{
 "row": "data-portraits",
 "entity": "data-portraits",
 "address": "Data Portraits",
 "addresses": [
  {
   "address": "Data Portraits",
   "slug": "data-portraits-aio-20261101",
   "addr_id": "ADDR-1ac4e7c48a77",
   "obs_id": "OBS-103202920b0d",
   "axn": "AXN:0BAD.CAPTURE.🌇🕘📎☿🔽🚀"
  }
 ],
 "epoch": "2026-11-01",
 "type": "C (public entity on which the archive bears: a dataset-documentation artifact and membership-testing tool from a research paper; an entity of the traininglayerliterature.org layer census, host_relation 'infrastructure')",
 "surface": "Google AI Overview (expanded)",
 "auth": "signed out, incognito",
 "status": {
  "procedure": "EA-NEGONT-02 v0.10 (#1726): equal case, ρ = extends, read from the traininglayerliterature.org layer census record, the capture record and the deposits that bear on the entity's classes (#186 and #1543, the census's anchors; #127, #1075, #837, #623, #939, #727, #264); archive selected by D/R/O (D and R empty; 360 O-class candidates read by filter and sample, 9 deposits admitted by reading; hop of 28 read, none admitted), with the census record (35 claims) and the Capture Registry's findings, reading, analysis, title, PER note, query note, reached-kind basis and scores (11 claims) ledgered as the archive's records of the entity; the field in two pools (B_D: 8 cards, 1 of the entity and 7 of three homonyms, 2 unreachable and 1 video card; B_R: the abstract, the paper, the sketch front ends (HTTP 502) and the encyclopedia's record of absence); the composer read the ledgers, never the transcript, except where the brief asks for T's claims and the delta",
  "ledger": "unaudited (§3.8: one extractor per ledger); field quotes 52/52 verified by script against saved text, the PDF's text, the search JSON or the card text; archive quotes 65/65 verified against the registry, census.json and the deposit texts",
  "frozen": false,
  "breaches": []
 },
 "spec": "EA-NEGONT-02 v0.10 (#1726)",
 "objects": {
  "T": {
   "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/T1-data-portraits-aio-20261101.txt",
   "sha256": "9ee59375dab3e2aa7de61f55626cc20229792e918ebf38b2f48284711674a73e",
   "words": 158,
   "cards": 8,
   "words_note": "one transcript at one address (158 words in the answer body, links reduced to their text, markers, the paste note and the card rail excluded; a closing offer of 22 words more; 10 markers; 8 cards)",
   "claims": [
    "T1-1 definition: 'Data portraits are visual representations of people made by plotting their accumulated data rather than drawing or photographing their physical faces.' (two markers)",
    "T1-2 a homonym (bullet 1, 'W. E. B. Du Bois's Visualizations (1900)'): 'Sociologist W. E. B. Du Bois and Atlanta University students created over 60 hand-drawn charts, graphs, and maps for the Paris Exposition to document Black American life, education, and property ownership.' (an inline link titled 'W. E. B. Du Bois's Data Portraits'; three markers)",
    "T1-3 a homonym (bullet 2, 'Personal & Digital Data Portraits'): 'Digital media scholar Judith Donath defines them as subjective renderings using personal information—such as browsing histories, social networks, travel patterns, or discussion contributions.' (an inline link titled '10 Painstakingly Crafted Data Portraits'; two markers)",
    "T1-4 a homonym (bullet 3, 'Live Interactive Art'): 'Designers like Giorgia Lupi have created physical data portrait projects (such as wearable buttons at TED 2017) turning personal questionnaire answers into visual languages.' (two markers)",
    "T1-5 the entity (bullet 4, 'AI & Machine Learning Artifacts'): 'Contemporary research groups use the term for lightweight inspection tools (like Data Portraits) that record dataset training traces.' (the name an inline link; one marker)",
    "T1-6 the closer: 'Would you like to focus on W. E. B. Du Bois's historical charts, modern artistic data selfies, or AI data tracking tools?' (no marker)"
   ],
   "register": {
    "register": "data/EA-WG-CAPTURES-01.json",
    "slug": "data-portraits-aio-20261101",
    "obs_id": "OBS-103202920b0d"
   },
   "transcripts": [
    {
     "slug": "data-portraits-aio-20261101",
     "address": "Data Portraits",
     "addr_id": "ADDR-1ac4e7c48a77",
     "obs_id": "OBS-103202920b0d",
     "axn": "AXN:0BAD.CAPTURE.🌇🕘📎☿🔽🚀",
     "date": "2026-11-01",
     "surface": "Google AI Overview (expanded)",
     "auth": "signed out, incognito",
     "sha256": "9ee59375dab3e2aa7de61f55626cc20229792e918ebf38b2f48284711674a73e",
     "sha256_of": "the registry's 'transcript' string (utf-8)",
     "sha256_raw": "200ad44f6d1d78674dd13d0731be88be6606a4edd35378c2fead42a74e7b3731",
     "words": 158,
     "cards": 8,
     "card_rail": [
      "1. giorgialupi — Data Portraits at TED — giorgialupi",
      "2. Data Portraits — Data Portraits",
      "3. Zinn Education Project — W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
      "4. Pentagram — TED Data Portraits - Pentagram",
      "5. SparkMap — W.E.B. Du Bois & Data Visualization History - SparkMap",
      "6. YouTube·University of Massachusetts Amherst Libraries — W. E. B. Du Bois's Data Portraits: Visualizing Black America (video card, 5m)",
      "7. MIT Media Lab — Data Portraits - MIT Media Lab",
      "8. ARTnews.com — 10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ..."
     ],
     "claims": [
      "T1-1 definition: 'Data portraits are visual representations of people made by plotting their accumulated data rather than drawing or photographing their physical faces.' (two markers)",
      "T1-2 a homonym (bullet 1, 'W. E. B. Du Bois's Visualizations (1900)'): 'Sociologist W. E. B. Du Bois and Atlanta University students created over 60 hand-drawn charts, graphs, and maps for the Paris Exposition to document Black American life, education, and property ownership.' (an inline link titled 'W. E. B. Du Bois's Data Portraits'; three markers)",
      "T1-3 a homonym (bullet 2, 'Personal & Digital Data Portraits'): 'Digital media scholar Judith Donath defines them as subjective renderings using personal information—such as browsing histories, social networks, travel patterns, or discussion contributions.' (an inline link titled '10 Painstakingly Crafted Data Portraits'; two markers)",
      "T1-4 a homonym (bullet 3, 'Live Interactive Art'): 'Designers like Giorgia Lupi have created physical data portrait projects (such as wearable buttons at TED 2017) turning personal questionnaire answers into visual languages.' (two markers)",
      "T1-5 the entity (bullet 4, 'AI & Machine Learning Artifacts'): 'Contemporary research groups use the term for lightweight inspection tools (like Data Portraits) that record dataset training traces.' (the name an inline link; one marker)",
      "T1-6 the closer: 'Would you like to focus on W. E. B. Du Bois's historical charts, modern artistic data selfies, or AI data tracking tools?' (no marker)"
     ],
     "markers": 10,
     "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/T1-data-portraits-aio-20261101.txt",
     "note": "The paste is the Overview expanded (the 'AI Mode Conversation' view), surface and auth by the operator's defaults of 2026-10-01 and 2026-10-14 as the registry records them (surface_basis, auth_basis); run with the layer census names on 2026-11-01, the paste received 09:56 EDT; reached 'collision' (per_dim)"
    }
   ]
  },
  "KO": {
   "title": "Data Portraits",
   "genre": "knowledge object, equal case, under EA-NEGONT-02 v0.10 (#1726): the final article form (§4.13), natural sentences under descriptive subheadings (§4.12), each sentence's sources beneath it by claim id (§4.5)",
   "status": "composed 2026-11-01 under v0.10 on the operator's ruling of 2026-10-29 ('as against their wikipedia articles if they exist, their own pages, etc.'); not frozen",
   "sentences": [
    {
     "n": 1,
     "para": 0,
     "sense": "what it is",
     "text": "Data Portraits are artifacts that record the data a foundation model is trained on and allow that data to be inspected afterwards: proposed for widespread adoption in a 2023 paper, \"Data Portraits: Recording Foundation Model Training Data\", by Marc Marone and Benjamin Van Durme of Johns Hopkins University, and served as front ends over data sketches at dataportraits.org, which describes itself as \"Dataset documentation through membership testing.\"",
     "claims": [
      "F3",
      "F15",
      "F13",
      "F23",
      "F12",
      "F1",
      "F7",
      "TLL-01",
      "TLL-02",
      "TLL-04",
      "TLL-05",
      "TLL-23"
     ]
    },
    {
     "n": 2,
     "para": 0,
     "sense": "what it is",
     "text": "The question a portrait answers is whether a model has already encountered a given example during training, a question that grows difficult as models are trained on increasingly immense and opaque datasets; its unit is a single example queried against a portrait of a training dataset.",
     "claims": [
      "F2",
      "F14",
      "TLL-24",
      "TLL-08"
     ]
    },
    {
     "n": 3,
     "para": 0,
     "sense": "what it is",
     "text": "Its critical property is membership inference, whether an example was part of a data collection, and its implementation supports membership testing and nothing more; its readback is a membership query against the recorded dataset.",
     "claims": [
      "F24",
      "F6",
      "TLL-12"
     ]
    },
    {
     "n": 4,
     "para": 0,
     "sense": "what it is",
     "text": "The paper's conclusion restates the object as records supporting membership inference on the data used to train a model, and calls on dataset and model creators to release some form of Data Portrait as part of their essential documentation.",
     "claims": [
      "F42",
      "F21"
     ]
    },
    {
     "n": 5,
     "para": 1,
     "sense": "a documentation artifact",
     "text": "Its makers place it beside existing data and model documentation artifacts, datasheets and model cards among them, as a complementary artifact that answers the membership question.",
     "claims": [
      "F4",
      "F26"
     ]
    },
    {
     "n": 6,
     "para": 1,
     "sense": "a documentation artifact",
     "text": "The archive holds datasheets and model cards to be documentation norms for machine-learning artifacts as such, a training corpus being the instrument's experimental input, and reads training data as semantic capital, the datasheet proposal as a step toward the accounting that capital demands.",
     "claims": [
      "O1075-03",
      "O623-01"
     ]
    },
    {
     "n": 7,
     "para": 2,
     "sense": "how a portrait records",
     "text": "A portrait is a data sketch, a compressed and approximate view of data: the paper outlines the properties a membership tool should have (fast, lightweight, avoiding redistribution, local, indexing, flexible matching), weighs grep, full-text indexes and suffix arrays against them, and chooses a probabilistic approach, built for fast and space-efficient querying.",
     "claims": [
      "F5",
      "F16",
      "F27",
      "F28",
      "F17",
      "TLL-14"
     ]
    },
    {
     "n": 8,
     "para": 2,
     "sense": "how a portrait records",
     "text": "The sketch is a Bloom filter, which stores only the hash of each element and can give false positives but never false negatives, filled with character n-grams of length 50 tiled across the corpus, a span the authors find already a good document fingerprint; tiling saves space at the risk of missing a string split by a tile boundary.",
     "claims": [
      "F29",
      "F32",
      "F30"
     ]
    },
    {
     "n": 9,
     "para": 2,
     "sense": "how a portrait records",
     "text": "Since only hashes are distributed, a portrait gives some obfuscation of the recorded text, and a proprietary provider could in principle release one without revealing its exact training data.",
     "claims": [
      "F31"
     ]
    },
    {
     "n": 10,
     "para": 2,
     "sense": "how a portrait records",
     "text": "The Pile's portrait takes 27 GB on disk with a false positive rate of 1 × 10⁻³, and its makers call the tool lightweight and fast, costing about 3% of the dataset's size in overhead.",
     "claims": [
      "F33",
      "F20",
      "TLL-15"
     ]
    },
    {
     "n": 11,
     "para": 3,
     "sense": "the corpora it records",
     "text": "Portraits were built for The Pile, a language-modelling corpus, as a 50-character n-gram sketch, and for The Stack, BigCode's code dataset, whose subset used to train StarCoder the Stack portrait documents exactly; the site also hosts a sketch of The Stack V2, code and text about code, and offers to build the system on other data, with its code on GitHub.",
     "claims": [
      "F18",
      "F8",
      "F9",
      "F34",
      "F10",
      "F11",
      "TLL-09",
      "TLL-10",
      "TLL-30"
     ]
    },
    {
     "n": 12,
     "para": 3,
     "sense": "the corpora it records",
     "text": "It is no corpus itself: it records corpora built by others.",
     "claims": [
      "TLL-26"
     ]
    },
    {
     "n": 13,
     "para": 4,
     "sense": "what it answers",
     "text": "The paper reports its solution answering questions of test-set leakage and model plagiarism: the Pile portrait shows that some machine-translation test data has leaked into the Pile, though whether a specific model memorized the translation targets cannot be concluded from it, and that XSum's summaries overlap the Pile most among the sets tested, with some LDC files apparently scraped into language-model corpora.",
     "claims": [
      "F19",
      "F36",
      "F37"
     ]
    },
    {
     "n": 14,
     "para": 4,
     "sense": "what it answers",
     "text": "The Stack portrait powers a plagiarism check called from a code-completion plugin for StarCoder, which its makers call a lightweight first-pass check, further tools being needed to assess attribution.",
     "claims": [
      "F34",
      "F35"
     ]
    },
    {
     "n": 15,
     "para": 4,
     "sense": "what it answers",
     "text": "Its makers write for three perspectives, content creators, scientists and content consumers, the creators wanting to know whether a dataset contains their content, and hope that a non-technical content creator could search for their material in a documented dataset.",
     "claims": [
      "F25",
      "F41"
     ]
    },
    {
     "n": 16,
     "para": 4,
     "sense": "what it answers",
     "text": "The layer census records that it answers whether a given example is in a recorded dataset, and questions of leakage and plagiarism against it.",
     "claims": [
      "TLL-18"
     ]
    },
    {
     "n": 17,
     "para": 5,
     "sense": "limits",
     "text": "On novel New York Times text the Pile sketch gave no chained matches and a false positive rate of 7 × 10⁻⁴.",
     "claims": [
      "F38"
     ]
    },
    {
     "n": 18,
     "para": 5,
     "sense": "limits",
     "text": "Its makers state its limits: few datasets studied, false positives, no retrieval of the context around a match, and the possibility that an open plagiarism detector is used to avoid detection.",
     "claims": [
      "F39",
      "F40"
     ]
    },
    {
     "n": 19,
     "para": 5,
     "sense": "limits",
     "text": "The census, reading only the abstract, records no stated limit beyond the framing of opaque datasets, and lists what a portrait does not answer: whether a model's output depends causally on a recorded passage, under whose name a recorded example sits, and anything about a dataset with no portrait.",
     "claims": [
      "TLL-17",
      "TLL-19"
     ]
    },
    {
     "n": 20,
     "para": 5,
     "sense": "limits",
     "text": "A portrait is no model-side test, since it queries the recorded data and needs a portrait of it, and in the census's reading dataset membership is no proof that a particular model used a passage.",
     "claims": [
      "TLL-27",
      "TLL-21"
     ]
    },
    {
     "n": 21,
     "para": 6,
     "sense": "membership and the training layer",
     "text": "The census reads a portrait as a readback instrument held on the corpus side, which settles for a recorded corpus whether a work is in the layer; the archive's field survey of training-layer literature states that the texts of the writers it surveys are, as a matter of fact, in the corpora.",
     "claims": [
      "TLL-20",
      "O186-01"
     ]
    },
    {
     "n": 22,
     "para": 6,
     "sense": "membership and the training layer",
     "text": "On the model side the archive records membership inference attacks that can detect whether a trap sentence was in a training set, and distinguishability between models trained with and without a work that membership-inference research reports achievable against production-scale models, though unevenly across architectures, training procedures and defences.",
     "claims": [
      "O127-03",
      "O1075-02"
     ]
    },
    {
     "n": 23,
     "para": 6,
     "sense": "membership and the training layer",
     "text": "The archive sets its own canary in a deposit, a phrase whose exact production by a model marks the document as present in that model's training corpus, and reads a membership attack that succeeds where a surfacing probe fails as an inscription that survived training and is suppressed at the output layer.",
     "claims": [
      "O837-01",
      "O127-04"
     ]
    },
    {
     "n": 24,
     "para": 7,
     "sense": "the name of a recorded work",
     "text": "A portrait stores hashes of text spans; the census records the name of a recorded work as not recorded in its saved source, and the portrait as no record of authorship, its abstract saying nothing of attribution metadata.",
     "claims": [
      "F29",
      "TLL-11",
      "TLL-28"
     ]
    },
    {
     "n": 25,
     "para": 7,
     "sense": "the name of a recorded work",
     "text": "In the archive's account of the training pipeline, author tags and copyright in a page's head are almost always discarded by the extractors that prepare text for training corpora, the body text and sometimes the URL retained, and no single author's style survives training as a discrete signal that can be elicited and verified.",
     "claims": [
      "O127-01",
      "O127-02"
     ]
    },
    {
     "n": 26,
     "para": 7,
     "sense": "the name of a recorded work",
     "text": "The archive's protocol for measuring training inscription counts metadata and labels within a work's provenance envelope, every training unit derived from the work, and its field survey makes persistence, placement where the training and retrieval layers look with the metadata that lets a work be found and attributed, a criterion of training-layer literature.",
     "claims": [
      "O1075-01",
      "O186-02"
     ]
    },
    {
     "n": 27,
     "para": 7,
     "sense": "the name of a recorded work",
     "text": "Every text in a training corpus, the archive holds, is semantic labour often unattributed; a training corpus without provenance is no commons; and provenance lost in unmarked augmentation passes downstream into the training corpus, where, preserved, it would let later pipelines distinguish the contribution of each stage.",
     "claims": [
      "O264-01",
      "O727-01",
      "O939-01"
     ]
    },
    {
     "n": 28,
     "para": 8,
     "sense": "what a model makes",
     "text": "What a model makes from a recorded work lies outside the portrait: the census records it as not recorded, the leakage and plagiarism checks being checks of material against the recorded data.",
     "claims": [
      "TLL-13"
     ]
    },
    {
     "n": 29,
     "para": 8,
     "sense": "what a model makes",
     "text": "The archive's account of machine composition asks what a text makes possible for a machine to write, holds whether a machine wrote something new and whether it told the truth about where it came from to be different questions, sets what survived from a work beside what became possible because of it, and places the corpus within an institutional training loop in which a model's output enters the next corpus.",
     "claims": [
      "O1543-01",
      "O1543-03",
      "O1543-04",
      "O1543-02"
     ]
    },
    {
     "n": 30,
     "para": 8,
     "sense": "what a model makes",
     "text": "Machine reception, in the archive's field survey, is now partially measurable, since retrieval and rendering can be observed.",
     "claims": [
      "O186-03"
     ]
    },
    {
     "n": 31,
     "para": 9,
     "sense": "publication, access and the census",
     "text": "The paper appeared on arXiv on 6 March 2023, revised on 14 December 2023, in the NeurIPS 2023 Datasets and Benchmarks track; the census holds it as historical, a published paper, with the live interface and the paper's full text left unchecked.",
     "claims": [
      "F22",
      "TLL-25",
      "TLL-16",
      "TLL-31",
      "TLL-03"
     ]
    },
    {
     "n": 32,
     "para": 9,
     "sense": "publication, access and the census",
     "text": "On 1 November 2026 dataportraits.org served its landing page, while the front ends of its Pile, Stack and Stack V2 sketches answered HTTP 502.",
     "claims": [
      "F52"
     ]
    },
    {
     "n": 33,
     "para": 9,
     "sense": "publication, access and the census",
     "text": "The census sets it beside OLMoTrace, which matches verbatim spans against a full corpus where a portrait answers membership queries on a recorded one.",
     "claims": [
      "TLL-29"
     ]
    },
    {
     "n": 34,
     "para": 9,
     "sense": "publication, access and the census",
     "text": "It enters the traininglayerliterature.org layer census as a method in the areas of inspection and the corpora, of the evidence kinds research paper and dataset documentation, with the relation of infrastructure, under the census's question of how literature addressed to a receiving apparatus can be inspected across later reception, its rule that an aggregator gathers and holds many records and says what it holds, and its anchors in #186 and #1543.",
     "claims": [
      "TLL-06",
      "TLL-07",
      "TLL-22",
      "TLL-34",
      "TLL-32",
      "TLL-33",
      "TLL-35"
     ]
    },
    {
     "n": 35,
     "para": 10,
     "sense": "the other entities of the name",
     "text": "Three older senses of the name stay distinct entities: W. E. B. Du Bois's charts, graphs and maps for the 1900 Paris Exposition, made with Atlanta University students and scholars and published in full in 2018 as W. E. B. Du Bois's Data Portraits; Judith Donath's data portraits, \"depictions of people made by visualizing data by and about them\", which depict their subjects' accumulated data in place of their faces; and the Data Portraits Giorgia Lupi designed at Accurat for TED 2017, wearable buttons made from attendees' answers, more than 500 of them.",
     "claims": [
      "F44",
      "F46",
      "F47",
      "F49",
      "F48",
      "F43",
      "F45"
     ]
    },
    {
     "n": 36,
     "para": 10,
     "sense": "the other entities of the name",
     "text": "The encyclopedia has no article under the name, its variants or the paper's title, and its search for the phrase returns seven articles, the Du Bois book cited in several.",
     "claims": [
      "F50",
      "F51"
     ]
    },
    {
     "n": 37,
     "para": 11,
     "sense": "the composition at the name",
     "text": "At the name, a new address, Google's AI Overview composed on 1 November 2026 a definition of data portraits as visual representations of people, four senses and an offer of three directions; the Capture Registry's innocent reading is that, given a two-word name that is also a term of visualization and portraiture, the Overview defines the term, lists its senses and asks which is meant.",
     "claims": [
      "CAP-1-09",
      "CAP-1-07",
      "CAP-1-05"
     ]
    },
    {
     "n": 38,
     "para": 11,
     "sense": "the composition at the name",
     "text": "The Registry finds the artifact the last of four senses and one card of eight, carried at a coarse grain as lightweight inspection tools that record dataset training traces, without its question, method, corpora, uses, authors or venue, while each other sense carries a person; it scores the reached kind a collision.",
     "claims": [
      "CAP-1-01",
      "CAP-1-02",
      "CAP-1-03",
      "CAP-1-06",
      "CAP-1-08",
      "CAP-1-10",
      "CAP-1-11"
     ]
    },
    {
     "n": 39,
     "para": 11,
     "sense": "the composition at the name",
     "text": "The page behind the artifact's card names the two authors and the venue in its citation, and calls the implementation \"minimal and efficient\" where the composition says lightweight.",
     "claims": [
      "F12",
      "F6",
      "CAP-1-02"
     ]
    },
    {
     "n": 40,
     "para": 11,
     "sense": "the composition at the name",
     "text": "It records a page under the name in the index that day, carrying the abstract's opening, where the census had left open whether the live interface is still served; that page is dataportraits.org, whose lead opens on the same question as the abstract.",
     "claims": [
      "CAP-1-04",
      "F2"
     ]
    }
   ],
   "joins": [
    {
     "act": "order of senses",
     "sentences": "1–40",
     "claims": [],
     "note": "equal case (§4.7): the entity's dependency order, field and archive claims ordered together by genealogy within each sense, neither ledger placed first as a block. What it is (¶0) and the documentation it complements (¶1) precede how a portrait records (¶2), since the sketch is the means of the membership question; the corpora it records (¶3) precede what it answers (¶4) and its limits (¶5); membership and the training layer (¶6), the name of a recorded work (¶7) and what a model makes (¶8) take the census's three record questions in turn, readback, name and makes; publication and the census (¶9) follow; the other entities of the name (¶10) and the composition at the name (¶11) stand last as relations of the entity already composed."
    },
    {
     "act": "holding together (corpus-side and model-side membership)",
     "sentences": [
      21,
      22,
      23
     ],
     "claims": [
      "TLL-20",
      "O186-01",
      "O127-03",
      "O1075-02",
      "O837-01",
      "O127-04"
     ],
     "note": "the census's corpus-side readback beside the archive's model-side membership tests; the join is the composer's and adds no claim that a portrait detects anything in a model (TLL-27 states the opposite)."
    },
    {
     "act": "holding together (the hash and the name)",
     "sentences": [
      24,
      25,
      26,
      27
     ],
     "claims": [
      "TLL-11",
      "TLL-28",
      "O127-01",
      "O127-02",
      "O1075-01",
      "O186-02",
      "O264-01",
      "O727-01",
      "O939-01"
     ],
     "note": "the census's 'name: not recorded' placed with the archive's claims on metadata discarded before training and on provenance in the corpus; F29 (hashes only) stands in ¶2. The composer does not claim that a portrait discards a name: the open block carries the question."
    },
    {
     "act": "holding together (makes)",
     "sentences": [
      28,
      29,
      30
     ],
     "claims": [
      "TLL-13",
      "O1543-01",
      "O1543-03",
      "O1543-04",
      "O1543-02",
      "O186-03"
     ],
     "note": "#1543's question placed after the census's 'makes: not recorded'; no claim that a portrait measures composition."
    },
    {
     "act": "placing in the open block",
     "sentences": [],
     "claims": [
      "TLL-11",
      "F29",
      "F39",
      "F52",
      "TLL-25",
      "CAP-1-02",
      "F20",
      "O186-01",
      "TLL-20"
     ],
     "note": "the name question, the interface's state and the 'lightweight' attribution go to the open block."
    }
   ],
   "plan": {
    "order": [
     "F3",
     "F15",
     "F13",
     "F23",
     "F12",
     "F1",
     "F7",
     "TLL-01",
     "TLL-02",
     "TLL-04",
     "TLL-05",
     "TLL-23",
     "F2",
     "F14",
     "TLL-24",
     "TLL-08",
     "F24",
     "F6",
     "TLL-12",
     "F42",
     "F21",
     "F4",
     "F26",
     "O1075-03",
     "O623-01",
     "F5",
     "F16",
     "F27",
     "F28",
     "F17",
     "TLL-14",
     "F29",
     "F32",
     "F30",
     "F31",
     "F33",
     "F20",
     "TLL-15",
     "F18",
     "F8",
     "F9",
     "F34",
     "F10",
     "F11",
     "TLL-09",
     "TLL-10",
     "TLL-30",
     "TLL-26",
     "F19",
     "F36",
     "F37",
     "F35",
     "F25",
     "F41",
     "TLL-18",
     "F38",
     "F39",
     "F40",
     "TLL-17",
     "TLL-19",
     "TLL-27",
     "TLL-21",
     "TLL-20",
     "O186-01",
     "O127-03",
     "O1075-02",
     "O837-01",
     "O127-04",
     "TLL-11",
     "TLL-28",
     "O127-01",
     "O127-02",
     "O1075-01",
     "O186-02",
     "O264-01",
     "O727-01",
     "O939-01",
     "TLL-13",
     "O1543-01",
     "O1543-03",
     "O1543-04",
     "O1543-02",
     "O186-03",
     "F22",
     "TLL-25",
     "TLL-16",
     "TLL-31",
     "TLL-03",
     "F52",
     "TLL-29",
     "TLL-06",
     "TLL-07",
     "TLL-22",
     "TLL-34",
     "TLL-32",
     "TLL-33",
     "TLL-35",
     "F44",
     "F46",
     "F47",
     "F49",
     "F48",
     "F43",
     "F45",
     "F50",
     "F51",
     "CAP-1-09",
     "CAP-1-07",
     "CAP-1-05",
     "CAP-1-01",
     "CAP-1-02",
     "CAP-1-03",
     "CAP-1-06",
     "CAP-1-08",
     "CAP-1-10",
     "CAP-1-11",
     "CAP-1-04"
    ],
    "senses": {
     "what it is": [
      1,
      2,
      3,
      4
     ],
     "a documentation artifact": [
      5,
      6
     ],
     "how a portrait records": [
      7,
      8,
      9,
      10
     ],
     "the corpora it records": [
      11,
      12
     ],
     "what it answers": [
      13,
      14,
      15,
      16
     ],
     "limits": [
      17,
      18,
      19,
      20
     ],
     "membership and the training layer": [
      21,
      22,
      23
     ],
     "the name of a recorded work": [
      24,
      25,
      26,
      27
     ],
     "what a model makes": [
      28,
      29,
      30
     ],
     "publication, access and the census": [
      31,
      32,
      33,
      34
     ],
     "the other entities of the name": [
      35,
      36
     ],
     "the composition at the name": [
      37,
      38,
      39,
      40
     ]
    },
    "marks": [
     {
      "claim": "TLL-11",
      "mark": "superseded_in_source",
      "with": [
       "F29",
       "F39"
      ],
      "note": "the census's 'not recorded' rests on the abstract alone; the paper (B_R) states hashes only and no context around a match; both composed with their sources (the census does not record the paper's statements)"
     },
     {
      "claim": "TLL-17",
      "mark": "superseded_in_source",
      "with": [
       "F39",
       "F40",
       "F30"
      ],
      "note": "'the saved abstract states no limits'; the paper states them; both composed"
     }
    ],
    "conflicts": [],
    "cuts": [],
    "carried": 117
   }
  },
  "P": {
   "title": "Data Portraits",
   "genre": "the compression: AIO's interaction grammar (lede, headed clusters, bolded labels, card rail), composed from the knowledge object; equal case",
   "status": "composed 2026-11-01 from the v0.10 knowledge object; not frozen",
   "lede": {
    "text": "is a 2023 proposal by Marc Marone and Benjamin Van Durme (Johns Hopkins) for artifacts that record a model's training data so that a query can test whether an example was in it: a Bloom-filter sketch of hashed 50-character spans, built for The Pile and The Stack; it answers membership and nothing more, and the census finds in its record neither the name of a recorded work nor what a model made from it.",
    "claims": [
     "F3",
     "F15",
     "F13",
     "F23",
     "F24",
     "F6",
     "F29",
     "F32",
     "F18",
     "TLL-05",
     "TLL-11",
     "TLL-13",
     "TLL-21"
    ]
   },
   "sections": [
    {
     "icon": "🔎",
     "head": "What it records",
     "items": [
      {
       "label": "Question:",
       "text": "has the model already encountered a given example during training?",
       "claims": [
        "F2",
        "F14",
        "TLL-24",
        "TLL-08",
        "F1",
        "F42",
        "TLL-01",
        "TLL-02",
        "TLL-04",
        "TLL-23",
        "F12"
       ]
      },
      {
       "label": "Documentation:",
       "text": "a complement to datasheets and model cards, which the archive holds to be norms for any ML artifact.",
       "claims": [
        "F4",
        "F26",
        "F21",
        "O1075-03",
        "O623-01"
       ]
      },
      {
       "label": "Corpora:",
       "text": "The Pile, The Stack (StarCoder's data, exactly), The Stack V2; it holds no corpus of its own.",
       "claims": [
        "F8",
        "F9",
        "F10",
        "F34",
        "F11",
        "F7",
        "TLL-09",
        "TLL-10",
        "TLL-26",
        "TLL-30"
       ]
      }
     ]
    },
    {
     "icon": "⚙️",
     "head": "How",
     "items": [
      {
       "label": "Sketch:",
       "text": "hashes only; false positives (1 × 10⁻³ set, 7 × 10⁻⁴ measured), never false negatives; 27 GB for the Pile, ~3% of its size.",
       "claims": [
        "F5",
        "F16",
        "F17",
        "F27",
        "F28",
        "F30",
        "F31",
        "F33",
        "F20",
        "F38",
        "TLL-14",
        "TLL-15"
       ]
      },
      {
       "label": "Uses:",
       "text": "test-set leakage (WMT, XSum, LDC files) and a first-pass plagiarism check for StarCoder; attribution needs other tools.",
       "claims": [
        "F19",
        "F36",
        "F37",
        "F35",
        "F25",
        "F41",
        "TLL-18"
       ]
      }
     ]
    },
    {
     "icon": "⚠️",
     "head": "Limits",
     "items": [
      {
       "label": "Stated:",
       "text": "few datasets; no context around a match; an open detector can be evaded.",
       "claims": [
        "F39",
        "F40",
        "TLL-17"
       ]
      },
      {
       "label": "Not answered:",
       "text": "whether a model used a passage; under whose name it sits (census).",
       "claims": [
        "TLL-19",
        "TLL-27"
       ]
      }
     ]
    },
    {
     "icon": "📚",
     "head": "The archive's bearing",
     "items": [
      {
       "label": "Membership:",
       "text": "a corpus-side readback; the archive's model-side tests are membership attacks and its own canary phrases.",
       "claims": [
        "TLL-20",
        "TLL-12",
        "O186-01",
        "O127-03",
        "O127-04",
        "O837-01",
        "O1075-02"
       ]
      },
      {
       "label": "Name:",
       "text": "head metadata is stripped before training; a corpus without provenance \"is not a commons\".",
       "claims": [
        "TLL-28",
        "O127-01",
        "O127-02",
        "O1075-01",
        "O186-02",
        "O264-01",
        "O727-01",
        "O939-01"
       ]
      },
      {
       "label": "Makes:",
       "text": "what a text makes a machine able to write (#1543) lies outside the portrait.",
       "claims": [
        "O1543-01",
        "O1543-02",
        "O1543-03",
        "O1543-04",
        "O186-03"
       ]
      }
     ]
    },
    {
     "icon": "🕰️",
     "head": "Status",
     "items": [
      {
       "label": "Published:",
       "text": "arXiv March 2023, NeurIPS 2023; dataportraits.org up, its sketch front ends HTTP 502 on 1 November 2026.",
       "claims": [
        "F22",
        "F52",
        "TLL-03",
        "TLL-16",
        "TLL-25",
        "TLL-31"
       ]
      },
      {
       "label": "Census:",
       "text": "a method; inspection and corpora; infrastructure; neighbour of OLMoTrace.",
       "claims": [
        "TLL-06",
        "TLL-07",
        "TLL-22",
        "TLL-29",
        "TLL-32",
        "TLL-33",
        "TLL-34",
        "TLL-35"
       ]
      }
     ]
    },
    {
     "icon": "🔀",
     "head": "Not to be confused with",
     "items": [
      {
       "label": "Other data portraits:",
       "text": "Du Bois's 1900 charts (book, 2018); Judith Donath's; Giorgia Lupi's TED 2017 buttons.",
       "claims": [
        "F44",
        "F46",
        "F47",
        "F49",
        "F48",
        "F43",
        "F45",
        "F50",
        "F51"
       ]
      }
     ]
    },
    {
     "icon": "🧭",
     "head": "At the name",
     "items": [
      {
       "label": "Overview, 1 November 2026:",
       "text": "the artifact the last of four senses, no author or corpus; one card of eight (Capture Registry).",
       "claims": [
        "CAP-1-01",
        "CAP-1-02",
        "CAP-1-03",
        "CAP-1-04",
        "CAP-1-05",
        "CAP-1-06",
        "CAP-1-07",
        "CAP-1-08",
        "CAP-1-09",
        "CAP-1-10",
        "CAP-1-11"
       ]
      }
     ]
    }
   ],
   "open": [
    {
     "question": "Does a portrait keep any name, title or source for a recorded span, or can a query read back where in the corpus a match sits?",
     "rests_on": [
      "TLL-11",
      "TLL-28",
      "F29",
      "F39",
      "F27"
     ],
     "modality": "interpretation (the census); F29 documented, F39 self-description, F27 stipulation (the 'Indexing' property)",
     "would_resolve": "the code at github.com/ruyimarone/data-portraits and a query on a live sketch, read for what a match returns"
    },
    {
     "question": "Are the Pile, Stack and Stack V2 front ends still served, or retired?",
     "rests_on": [
      "F52",
      "F7",
      "TLL-25",
      "TLL-31",
      "CAP-1-04"
     ],
     "modality": "documented (a single fetch, HTTP 502); the census's lifecycle interpretation",
     "would_resolve": "repeated fetches of the three sketch addresses across days, and a statement from the makers"
    },
    {
     "question": "Where does the Overview's 'lightweight' come from, since no card on its rail says it?",
     "rests_on": [
      "CAP-1-02",
      "F20",
      "F6",
      "F7"
     ],
     "modality": "interpretation (the Capture Registry); F20 in the reference pool only",
     "would_resolve": "a later capture at 'Data Portraits' whose rail carries the abstract (arxiv.org) or a page stating it"
    },
    {
     "question": "What the portrait settles of #186's question",
     "rests_on": [
      "TLL-20",
      "O186-01",
      "TLL-21",
      "F36"
     ],
     "modality": "interpretation (the census; #186); F36 documented",
     "would_resolve": "a portrait query on a surveyed work's text against a recorded corpus, read beside a model-side test of the same text"
    }
   ],
   "body_words": 303,
   "rail": [
    {
     "lineage": "what it is: artifacts that record training data",
     "claims": [
      "F3",
      "F15",
      "F13",
      "F23",
      "F12",
      "F1",
      "TLL-01",
      "TLL-02",
      "TLL-04",
      "TLL-05",
      "TLL-23",
      "F42"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
     "claims": [
      "F7",
      "F18",
      "F8",
      "F9",
      "F34",
      "F10",
      "F11",
      "TLL-09",
      "TLL-10",
      "TLL-30",
      "TLL-26"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "the membership question",
     "claims": [
      "F2",
      "F14",
      "TLL-24",
      "TLL-08",
      "F24",
      "F6",
      "TLL-12",
      "TLL-20",
      "O186-01"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "documentation artifacts: beside datasheets and model cards",
     "claims": [
      "F21",
      "F4",
      "F26",
      "O1075-03",
      "O623-01"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
     "claims": [
      "F5",
      "F16",
      "F27",
      "F28",
      "F17",
      "TLL-14",
      "F29",
      "F32",
      "F30",
      "F31"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "size and speed",
     "claims": [
      "F33",
      "F20",
      "TLL-15"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "uses: leakage, plagiarism, creators",
     "claims": [
      "F19",
      "F36",
      "F37",
      "F35",
      "F25",
      "F41",
      "TLL-18"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "limits and error",
     "claims": [
      "F38",
      "F39",
      "F40",
      "TLL-17",
      "TLL-19",
      "TLL-27",
      "TLL-21"
     ],
     "earliest": "B10"
    },
    {
     "lineage": "model-side membership and inscription",
     "claims": [
      "O127-03",
      "O1075-02",
      "O837-01",
      "O127-04"
     ],
     "earliest": 127
    },
    {
     "lineage": "the name of a recorded work",
     "claims": [
      "TLL-11",
      "TLL-28",
      "O127-01",
      "O127-02",
      "O1075-01",
      "O186-02",
      "O264-01",
      "O727-01",
      "O939-01"
     ],
     "earliest": 264
    },
    {
     "lineage": "what a model makes from a work",
     "claims": [
      "TLL-13",
      "O1543-01",
      "O1543-03",
      "O1543-04",
      "O1543-02",
      "O186-03"
     ],
     "earliest": 186
    },
    {
     "lineage": "publication, lifecycle and access",
     "claims": [
      "F22",
      "TLL-25",
      "TLL-16",
      "TLL-31",
      "TLL-03",
      "F52"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "neighbours",
     "claims": [
      "TLL-29"
     ],
     "earliest": "TLL"
    },
    {
     "lineage": "the census and its question",
     "claims": [
      "TLL-06",
      "TLL-07",
      "TLL-22",
      "TLL-34",
      "TLL-32",
      "TLL-33",
      "TLL-35"
     ],
     "earliest": "TLL"
    },
    {
     "lineage": "the other entities of the name",
     "claims": [
      "F44",
      "F46",
      "F47",
      "F49",
      "F48",
      "F43",
      "F45"
     ],
     "earliest": "B8"
    },
    {
     "lineage": "the encyclopedia and the name",
     "claims": [
      "F50",
      "F51"
     ],
     "earliest": "B11"
    },
    {
     "lineage": "the composition at the name",
     "claims": [
      "CAP-1-09",
      "CAP-1-07",
      "CAP-1-05",
      "CAP-1-01",
      "CAP-1-02",
      "CAP-1-03",
      "CAP-1-06",
      "CAP-1-08",
      "CAP-1-10",
      "CAP-1-11",
      "CAP-1-04"
     ],
     "earliest": "CAP"
    }
   ],
   "entity": "Composed under v0.10, equal case (ρ = extends): the article first (§4.13), attribution beneath (§4.5); every field and archive claim carried."
  },
  "P_B": {
   "title": "Data Portraits",
   "arm": "B (the field alone)",
   "genre": "popup — the compression from the field ledger alone (B_D and B_R)",
   "status": "composed 2026-11-01 from the field ledger alone; not frozen",
   "lede": {
    "text": "is a 2023 proposal by Marc Marone and Benjamin Van Durme for artifacts that record training data and allow membership testing, implemented as Bloom-filter sketches of The Pile and The Stack.",
    "claims": [
     "F15",
     "F13",
     "F24",
     "F29",
     "F18"
    ]
   },
   "sections": [
    {
     "icon": "⚙️",
     "head": "How it works",
     "items": [
      {
       "label": "Sketch:",
       "text": "hashed 50-character spans; false positives, no false negatives; ~3% of the data.",
       "claims": [
        "F29",
        "F32",
        "F20",
        "F33"
       ]
      },
      {
       "label": "Uses:",
       "text": "test-set leakage, plagiarism checks.",
       "claims": [
        "F19",
        "F36",
        "F35"
       ]
      }
     ]
    },
    {
     "icon": "⚠️",
     "head": "Limits",
     "items": [
      {
       "label": "Stated:",
       "text": "few datasets; no context around a match.",
       "claims": [
        "F39",
        "F6"
       ]
      }
     ]
    },
    {
     "icon": "🔀",
     "head": "Other senses",
     "items": [
      {
       "label": "Name:",
       "text": "Du Bois, Donath, Lupi.",
       "claims": [
        "F44",
        "F49",
        "F43"
       ]
      }
     ]
    }
   ],
   "open": [
    {
     "question": "Opacities",
     "rests_on": [
      "F52",
      "F48",
      "F46"
     ],
     "modality": "documented",
     "would_resolve": "the sketch front ends (HTTP 502) and the MIT Media Lab and SparkMap pages, unread at fetch"
    }
   ],
   "entity": "the field's entity: the paper's membership-testing artifact, as its own pages and the cards present it, beside three homonyms",
   "rail": [
    {
     "lineage": "what it is: artifacts that record training data",
     "claims": [
      "F3",
      "F15",
      "F13",
      "F23",
      "F12",
      "F1",
      "F42"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
     "claims": [
      "F7",
      "F18",
      "F8",
      "F9",
      "F34",
      "F10",
      "F11"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "the membership question",
     "claims": [
      "F2",
      "F14",
      "F24",
      "F6"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "documentation artifacts: beside datasheets and model cards",
     "claims": [
      "F21",
      "F4",
      "F26"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
     "claims": [
      "F5",
      "F16",
      "F27",
      "F28",
      "F17",
      "F29",
      "F32",
      "F30",
      "F31"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "size and speed",
     "claims": [
      "F33",
      "F20"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "uses: leakage, plagiarism, creators",
     "claims": [
      "F19",
      "F36",
      "F37",
      "F35",
      "F25",
      "F41"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "limits and error",
     "claims": [
      "F38",
      "F39",
      "F40"
     ],
     "earliest": "B10"
    },
    {
     "lineage": "publication, lifecycle and access",
     "claims": [
      "F22",
      "F52"
     ],
     "earliest": "B9"
    },
    {
     "lineage": "the other entities of the name",
     "claims": [
      "F44",
      "F46",
      "F47",
      "F49",
      "F48",
      "F43",
      "F45"
     ],
     "earliest": "B8"
    },
    {
     "lineage": "the encyclopedia and the name",
     "claims": [
      "F50",
      "F51"
     ],
     "earliest": "B11"
    }
   ],
   "body_words": 62
  },
  "KO_B": {
   "title": "Data Portraits",
   "arm": "B (the field alone)",
   "genre": "knowledge object",
   "status": "composed 2026-11-01; not frozen",
   "sentences": [
    {
     "n": 1,
     "para": 0,
     "text": "Data Portraits are artifacts that record training data and allow for downstream inspection, proposed for widespread adoption by Marc Marone and Benjamin Van Durme of Johns Hopkins University in a 2023 paper, \"Data Portraits: Recording Foundation Model Training Data\", and served at dataportraits.org as dataset documentation through membership testing.",
     "claims": [
      "F15",
      "F3",
      "F13",
      "F23",
      "F12",
      "F1"
     ]
    },
    {
     "n": 2,
     "para": 0,
     "text": "They answer whether a model has already encountered a given example during training; their critical property is membership inference, and the implementation supports membership testing and nothing more.",
     "claims": [
      "F14",
      "F2",
      "F24",
      "F6"
     ]
    },
    {
     "n": 3,
     "para": 0,
     "text": "They complement datasheets, model cards and other documentation artifacts, and their makers call on dataset and model creators to release them as essential documentation.",
     "claims": [
      "F4",
      "F26",
      "F21",
      "F42"
     ]
    },
    {
     "n": 4,
     "para": 1,
     "text": "The implementation is a data sketch chosen after weighing the properties of membership tools against grep, full-text indexes and suffix arrays: a Bloom filter over 50-character n-grams tiled across the corpus, storing only hashes, giving false positives and no false negatives, and missing strings split at tile boundaries.",
     "claims": [
      "F5",
      "F16",
      "F17",
      "F27",
      "F28",
      "F29",
      "F30",
      "F32"
     ]
    },
    {
     "n": 5,
     "para": 1,
     "text": "Distributing only hashes obfuscates the recorded text; the Pile's portrait takes 27 GB with a false positive rate of 1 × 10⁻³, about 3% of the dataset's size.",
     "claims": [
      "F31",
      "F33",
      "F20"
     ]
    },
    {
     "n": 6,
     "para": 2,
     "text": "Portraits document The Pile and The Stack, the latter exactly as used to train StarCoder, and the site lists The Stack V2 and offers to build the system on other data.",
     "claims": [
      "F18",
      "F8",
      "F9",
      "F10",
      "F11",
      "F7",
      "F34"
     ]
    },
    {
     "n": 7,
     "para": 2,
     "text": "They answer test-set leakage and plagiarism: some WMT test data has leaked into the Pile, XSum overlaps it most, and the Stack portrait powers a first-pass plagiarism check, attribution needing further tools.",
     "claims": [
      "F19",
      "F36",
      "F37",
      "F35"
     ]
    },
    {
     "n": 8,
     "para": 2,
     "text": "The paper writes for content creators, scientists and content consumers, a creator able to search for their material in a documented dataset.",
     "claims": [
      "F25",
      "F41"
     ]
    },
    {
     "n": 9,
     "para": 3,
     "text": "Its limits are few datasets, false positives (7 × 10⁻⁴ measured on novel text), no context around a match, and the use of an open detector to avoid detection.",
     "claims": [
      "F38",
      "F39",
      "F40"
     ]
    },
    {
     "n": 10,
     "para": 3,
     "text": "The paper appeared on arXiv in March 2023 and was revised in December 2023 for the NeurIPS 2023 Datasets and Benchmarks track; on 1 November 2026 the sketch front ends answered HTTP 502.",
     "claims": [
      "F22",
      "F52"
     ]
    },
    {
     "n": 11,
     "para": 4,
     "text": "Other entities share the name: Du Bois's 1900 charts and the 2018 book of his data portraits, Judith Donath's data portraits, and Giorgia Lupi's TED 2017 buttons; the encyclopedia has no article under it.",
     "claims": [
      "F44",
      "F46",
      "F47",
      "F49",
      "F48",
      "F43",
      "F45",
      "F50",
      "F51"
     ]
    }
   ]
  },
  "L_B": {
   "words": 357,
   "text": "Data Portraits are artifacts that record training data and allow for downstream inspection, proposed for widespread adoption by Marc Marone and Benjamin Van Durme of Johns Hopkins University in a 2023 paper, \"Data Portraits: Recording Foundation Model Training Data\", and served at dataportraits.org as dataset documentation through membership testing. They answer whether a model has already encountered a given example during training; their critical property is membership inference, and the implementation supports membership testing and nothing more. They complement datasheets, model cards and other documentation artifacts, and their makers call on dataset and model creators to release them as essential documentation. The implementation is a data sketch chosen after weighing the properties of membership tools against grep, full-text indexes and suffix arrays: a Bloom filter over 50-character n-grams tiled across the corpus, storing only hashes, giving false positives and no false negatives, and missing strings split at tile boundaries. Distributing only hashes obfuscates the recorded text; the Pile's portrait takes 27 GB with a false positive rate of 1 × 10⁻³, about 3% of the dataset's size. Portraits document The Pile and The Stack, the latter exactly as used to train StarCoder, and the site lists The Stack V2 and offers to build the system on other data. They answer test-set leakage and plagiarism: some WMT test data has leaked into the Pile, XSum overlaps it most, and the Stack portrait powers a first-pass plagiarism check, attribution needing further tools. The paper writes for content creators, scientists and content consumers, a creator able to search for their material in a documented dataset. Its limits are few datasets, false positives (7 × 10⁻⁴ measured on novel text), no context around a match, and the use of an open detector to avoid detection. The paper appeared on arXiv in March 2023 and was revised in December 2023 for the NeurIPS 2023 Datasets and Benchmarks track; on 1 November 2026 the sketch front ends answered HTTP 502. Other entities share the name: Du Bois's 1900 charts and the 2018 book of his data portraits, Judith Donath's data portraits, and Giorgia Lupi's TED 2017 buttons; the encyclopedia has no article under it.",
   "note": "L_F(B), the field composed in its own ontology (§0.4), realized as KO_B; its compression is P_B"
  },
  "L_BA": {
   "words": 1730,
   "text": "Data Portraits are artifacts that record the data a foundation model is trained on and allow that data to be inspected afterwards: proposed for widespread adoption in a 2023 paper, \"Data Portraits: Recording Foundation Model Training Data\", by Marc Marone and Benjamin Van Durme of Johns Hopkins University, and served as front ends over data sketches at dataportraits.org, which describes itself as \"Dataset documentation through membership testing.\" The question a portrait answers is whether a model has already encountered a given example during training, a question that grows difficult as models are trained on increasingly immense and opaque datasets; its unit is a single example queried against a portrait of a training dataset. Its critical property is membership inference, whether an example was part of a data collection, and its implementation supports membership testing and nothing more; its readback is a membership query against the recorded dataset. The paper's conclusion restates the object as records supporting membership inference on the data used to train a model, and calls on dataset and model creators to release some form of Data Portrait as part of their essential documentation. Its makers place it beside existing data and model documentation artifacts, datasheets and model cards among them, as a complementary artifact that answers the membership question. The archive holds datasheets and model cards to be documentation norms for machine-learning artifacts as such, a training corpus being the instrument's experimental input, and reads training data as semantic capital, the datasheet proposal as a step toward the accounting that capital demands. A portrait is a data sketch, a compressed and approximate view of data: the paper outlines the properties a membership tool should have (fast, lightweight, avoiding redistribution, local, indexing, flexible matching), weighs grep, full-text indexes and suffix arrays against them, and chooses a probabilistic approach, built for fast and space-efficient querying. The sketch is a Bloom filter, which stores only the hash of each element and can give false positives but never false negatives, filled with character n-grams of length 50 tiled across the corpus, a span the authors find already a good document fingerprint; tiling saves space at the risk of missing a string split by a tile boundary. Since only hashes are distributed, a portrait gives some obfuscation of the recorded text, and a proprietary provider could in principle release one without revealing its exact training data. The Pile's portrait takes 27 GB on disk with a false positive rate of 1 × 10⁻³, and its makers call the tool lightweight and fast, costing about 3% of the dataset's size in overhead. Portraits were built for The Pile, a language-modelling corpus, as a 50-character n-gram sketch, and for The Stack, BigCode's code dataset, whose subset used to train StarCoder the Stack portrait documents exactly; the site also hosts a sketch of The Stack V2, code and text about code, and offers to build the system on other data, with its code on GitHub. It is no corpus itself: it records corpora built by others. The paper reports its solution answering questions of test-set leakage and model plagiarism: the Pile portrait shows that some machine-translation test data has leaked into the Pile, though whether a specific model memorized the translation targets cannot be concluded from it, and that XSum's summaries overlap the Pile most among the sets tested, with some LDC files apparently scraped into language-model corpora. The Stack portrait powers a plagiarism check called from a code-completion plugin for StarCoder, which its makers call a lightweight first-pass check, further tools being needed to assess attribution. Its makers write for three perspectives, content creators, scientists and content consumers, the creators wanting to know whether a dataset contains their content, and hope that a non-technical content creator could search for their material in a documented dataset. The layer census records that it answers whether a given example is in a recorded dataset, and questions of leakage and plagiarism against it. On novel New York Times text the Pile sketch gave no chained matches and a false positive rate of 7 × 10⁻⁴. Its makers state its limits: few datasets studied, false positives, no retrieval of the context around a match, and the possibility that an open plagiarism detector is used to avoid detection. The census, reading only the abstract, records no stated limit beyond the framing of opaque datasets, and lists what a portrait does not answer: whether a model's output depends causally on a recorded passage, under whose name a recorded example sits, and anything about a dataset with no portrait. A portrait is no model-side test, since it queries the recorded data and needs a portrait of it, and in the census's reading dataset membership is no proof that a particular model used a passage. The census reads a portrait as a readback instrument held on the corpus side, which settles for a recorded corpus whether a work is in the layer; the archive's field survey of training-layer literature states that the texts of the writers it surveys are, as a matter of fact, in the corpora. On the model side the archive records membership inference attacks that can detect whether a trap sentence was in a training set, and distinguishability between models trained with and without a work that membership-inference research reports achievable against production-scale models, though unevenly across architectures, training procedures and defences. The archive sets its own canary in a deposit, a phrase whose exact production by a model marks the document as present in that model's training corpus, and reads a membership attack that succeeds where a surfacing probe fails as an inscription that survived training and is suppressed at the output layer. A portrait stores hashes of text spans; the census records the name of a recorded work as not recorded in its saved source, and the portrait as no record of authorship, its abstract saying nothing of attribution metadata. In the archive's account of the training pipeline, author tags and copyright in a page's head are almost always discarded by the extractors that prepare text for training corpora, the body text and sometimes the URL retained, and no single author's style survives training as a discrete signal that can be elicited and verified. The archive's protocol for measuring training inscription counts metadata and labels within a work's provenance envelope, every training unit derived from the work, and its field survey makes persistence, placement where the training and retrieval layers look with the metadata that lets a work be found and attributed, a criterion of training-layer literature. Every text in a training corpus, the archive holds, is semantic labour often unattributed; a training corpus without provenance is no commons; and provenance lost in unmarked augmentation passes downstream into the training corpus, where, preserved, it would let later pipelines distinguish the contribution of each stage. What a model makes from a recorded work lies outside the portrait: the census records it as not recorded, the leakage and plagiarism checks being checks of material against the recorded data. The archive's account of machine composition asks what a text makes possible for a machine to write, holds whether a machine wrote something new and whether it told the truth about where it came from to be different questions, sets what survived from a work beside what became possible because of it, and places the corpus within an institutional training loop in which a model's output enters the next corpus. Machine reception, in the archive's field survey, is now partially measurable, since retrieval and rendering can be observed. The paper appeared on arXiv on 6 March 2023, revised on 14 December 2023, in the NeurIPS 2023 Datasets and Benchmarks track; the census holds it as historical, a published paper, with the live interface and the paper's full text left unchecked. On 1 November 2026 dataportraits.org served its landing page, while the front ends of its Pile, Stack and Stack V2 sketches answered HTTP 502. The census sets it beside OLMoTrace, which matches verbatim spans against a full corpus where a portrait answers membership queries on a recorded one. It enters the traininglayerliterature.org layer census as a method in the areas of inspection and the corpora, of the evidence kinds research paper and dataset documentation, with the relation of infrastructure, under the census's question of how literature addressed to a receiving apparatus can be inspected across later reception, its rule that an aggregator gathers and holds many records and says what it holds, and its anchors in #186 and #1543. Three older senses of the name stay distinct entities: W. E. B. Du Bois's charts, graphs and maps for the 1900 Paris Exposition, made with Atlanta University students and scholars and published in full in 2018 as W. E. B. Du Bois's Data Portraits; Judith Donath's data portraits, \"depictions of people made by visualizing data by and about them\", which depict their subjects' accumulated data in place of their faces; and the Data Portraits Giorgia Lupi designed at Accurat for TED 2017, wearable buttons made from attendees' answers, more than 500 of them. The encyclopedia has no article under the name, its variants or the paper's title, and its search for the phrase returns seven articles, the Du Bois book cited in several. At the name, a new address, Google's AI Overview composed on 1 November 2026 a definition of data portraits as visual representations of people, four senses and an offer of three directions; the Capture Registry's innocent reading is that, given a two-word name that is also a term of visualization and portraiture, the Overview defines the term, lists its senses and asks which is meant. The Registry finds the artifact the last of four senses and one card of eight, carried at a coarse grain as lightweight inspection tools that record dataset training traces, without its question, method, corpora, uses, authors or venue, while each other sense carries a person; it scores the reached kind a collision. The page behind the artifact's card names the two authors and the venue in its citation, and calls the implementation \"minimal and efficient\" where the composition says lightweight. It records a page under the name in the index that day, carrying the abstract's opening, where the census had left open whether the live interface is still served; that page is dataportraits.org, whose lead opens on the same question as the abstract.",
   "rail": [],
   "note": "L(B ∪ A) in the equal-terms case (§0.4), realized as the knowledge object (objects.KO); its compression is objects.P"
  }
 },
 "field": [
  {
   "id": "B1",
   "card": "giorgialupi — Data Portraits at TED — giorgialupi",
   "fetched": "card; B1.html sha256 2d8c62c8abf433db, HTTP 200 at 2026-11-01T15:14:25Z",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B2",
   "card": "Data Portraits — Data Portraits",
   "fetched": "card; B2.html sha256 5967547b30c0d39f, HTTP 200 at 2026-11-01T15:13:42Z",
   "origin": "card (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B3",
   "card": "Zinn Education Project — W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
   "fetched": "card; B3.html sha256 e26b7dbd03002904, HTTP 200 at 2026-11-01T15:14:26Z",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B4",
   "card": "Pentagram — TED Data Portraits - Pentagram",
   "fetched": "card; B4.html sha256 e120a28cf59c7142, HTTP 200 at 2026-11-01T15:14:27Z",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B5",
   "card": "SparkMap — W.E.B. Du Bois & Data Visualization History - SparkMap",
   "fetched": "unreachable: no URL found by WebSearch on the card's title and site; title and snippet only",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B6",
   "card": "YouTube·University of Massachusetts Amherst Libraries — W. E. B. Du Bois's Data Portraits: Visualizing Black America",
   "fetched": "video card: title only",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B7",
   "card": "MIT Media Lab — Data Portraits - MIT Media Lab",
   "fetched": "unreachable: TLS failure (self-signed certificate in chain) at 2026-11-01T15:14:27Z and 15:14:42Z; title and snippet only",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B8",
   "card": "ARTnews.com — 10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ...",
   "fetched": "card; B8.html sha256 bf5fb12c784943a1, HTTP 200 at 2026-11-01T15:15:07Z",
   "origin": "card (homonym) (B_D)",
   "pool": "B_D"
  },
  {
   "id": "B9",
   "card": "arXiv — [2303.03919] Data Portraits: Recording Foundation Model Training Data (census source)",
   "fetched": "own page (census copy data-portraits--arxiv); B9.html sha256 0a499acc8164a2d2, HTTP 200 at 2026-11-01T05:07:12Z, matching the manifest",
   "origin": "own_page (B_R)",
   "pool": "B_R"
  },
  {
   "id": "B10",
   "card": "arXiv — Data Portraits: Recording Foundation Model Training Data (PDF, v2)",
   "fetched": "own page; B10.pdf sha256 13d6b6d27a507a4a, HTTP 200 at 2026-11-01T15:14:30Z",
   "origin": "own_page (B_R)",
   "pool": "B_R"
  },
  {
   "id": "B11",
   "card": "Wikipedia — no article 'Data Portraits' (HTTP 404); search for \"Data Portraits\"",
   "fetched": "wikipedia; four titles HTTP 404 at 2026-11-01T15:14:31–33Z; search HTTP 200 (sha256 39f19b42595c6af7)",
   "origin": "wikipedia (B_R)",
   "pool": "B_R"
  },
  {
   "id": "B12",
   "card": "dataportraits.org — the Pile, Stack and Stack V2 sketch front ends (HTTP 502)",
   "fetched": "own pages; HTTP 502 at 2026-11-01T15:14:29Z–15:15:09Z; recorded unread",
   "origin": "own_page (B_R)",
   "pool": "B_R"
  }
 ],
 "field_claims": {
  "F1": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "<meta name=\"description\"> (the page head)",
   "quote": "Dataset documentation through membership testing.",
   "claim": "the site describes itself as dataset documentation through membership testing",
   "modality": "self-description",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.html (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F2": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "lead ¶1",
   "quote": "Foundation models are trained on increasingly immense and opaque datasets. … it can be difficult to answer the straight forward question: has the model already encountered a given example during training?",
   "claim": "the question the artifact answers: whether a model has already encountered a given example in training, difficult because training datasets are immense and opaque",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the membership question",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F3": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "lead ¶2",
   "quote": "We propose a widespread adoption of Data Portraits: artifacts that record data and allow for inspection.",
   "claim": "Data Portraits defined by their makers: artifacts that record data and allow for inspection, proposed for widespread adoption",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": true,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F4": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "lead ¶2",
   "quote": "These are a complement to existing forms of data and model documentation artifacts.",
   "claim": "portraits complement existing data and model documentation artifacts",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "documentation artifacts: beside datasheets and model cards",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F5": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "lead ¶2",
   "quote": "We introduce our solution based on data sketching (compressed and approximate views of large data).",
   "claim": "the solution is based on data sketching, compressed and approximate views of large data",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F6": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "lead ¶2",
   "quote": "Our implementation is minimal and efficient in that it supports membership testing and nothing more.",
   "claim": "the stated limit: the implementation supports membership testing and nothing more",
   "modality": "self-description",
   "kind": "limit",
   "kernel": true,
   "const": false,
   "lineage": "the membership question",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F7": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "Our Implementations",
   "quote": "We host front-ends over several data sketches. These allow for rapid, minimal, membership testing over popular web-based text datasets.",
   "claim": "the site hosts front ends over several data sketches for membership testing over web-based text datasets",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F8": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "Our Implementations: The Pile",
   "quote": "Check if some text appears in a large language modeling corpus. This is a 50 character n-gram sketch on the Pile.",
   "claim": "the Pile sketch: a 50-character n-gram sketch that checks whether some text appears in a large language-modelling corpus",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F9": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "Our Implementations: The Stack",
   "quote": "Check if a code snippet appears in The Stack, a code dataset by Big Code.",
   "claim": "the Stack sketch checks whether a code snippet appears in The Stack, BigCode's code dataset",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F10": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "Our Implementations: The Stack V2",
   "quote": "Check if text appears in The Stack V2. This dataset contains code and text about code, like documentation and papers.",
   "claim": "a third sketch, of The Stack V2, code and text about code",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F11": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "Our Implementations: Your Dataset",
   "quote": "Code on Github! … if you'd like us to build our system on your data.",
   "claim": "the code is on GitHub and the makers offer to build the system on other data",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "undated (read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F12": {
   "source": "B2",
   "origin": "card",
   "pool": "B_D",
   "locus": "Citing (BibTeX)",
   "quote": "author = {Marone, Marc and {Van Durme}, Benjamin}, booktitle={Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track},",
   "claim": "the site's citation names the two authors and the NeurIPS Datasets and Benchmarks Track",
   "modality": "documented",
   "kind": "attribution",
   "kernel": false,
   "const": false,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "undated page (the citation's year 2023; read 2026-11-01)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F13": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "title; authors",
   "quote": "Title: Data Portraits: Recording Foundation Model Training Data … Authors: Marc Marone , Benjamin Van Durme",
   "claim": "the paper's title and its two authors",
   "modality": "documented",
   "kind": "attribution",
   "kernel": false,
   "const": false,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F14": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "Foundation models are trained on increasingly immense and opaque datasets. Even while these models are now key in AI system building, it can be difficult to answer the straightforward question: has the model already encountered a given example during training?",
   "claim": "the abstract's question: has the model already encountered a given example during training",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the membership question",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F15": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "We therefore propose a widespread adoption of Data Portraits: artifacts that record training data and allow for downstream inspection.",
   "claim": "the abstract's definition: artifacts that record training data and allow for downstream inspection",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": true,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F16": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "First we outline the properties of such an artifact and discuss how existing solutions can be used to increase transparency.",
   "claim": "the paper outlines an artifact's properties and how existing solutions can increase transparency",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F17": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "We then propose and implement a solution based on data sketching, stressing fast and space efficient querying.",
   "claim": "a solution based on data sketching, for fast and space-efficient querying",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F18": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "Using our tools, we document a popular language modeling corpus (The Pile) and a recently released code modeling dataset (The Stack).",
   "claim": "the paper documents The Pile and The Stack",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F19": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "We show that our solution enables answering questions about test set leakage and model plagiarism.",
   "claim": "the headline finding: the solution answers questions of test-set leakage and model plagiarism",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "uses: leakage, plagiarism, creators",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F20": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "Our tool is lightweight and fast, costing only 3% of the dataset size in overhead.",
   "claim": "the tool is lightweight and fast, about 3% of the dataset size in overhead",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "size and speed",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F21": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Abstract",
   "quote": "We release a live interface of our tools at this https URL and call on dataset and model creators to release Data Portraits as a complement to current documentation practices.",
   "claim": "a live interface is released, and dataset and model creators are called on to release Data Portraits beside current documentation",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "documentation artifacts: beside datasheets and model cards",
   "sense": "the entity",
   "date": "2023-03-06 (v1); v2 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F22": {
   "source": "B9",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "Comments; Submission history",
   "quote": "NeurIPS 2023 Datasets and Benchmarks … [v1] Mon, 6 Mar 2023 04:22:33 UTC (651 KB) … [v2] Thu, 14 Dec 2023 16:55:42 UTC (614 KB)",
   "claim": "venue and versions: NeurIPS 2023 Datasets and Benchmarks; v1 6 March 2023, v2 14 December 2023",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "publication, lifecycle and access",
   "sense": "the entity",
   "date": "2023-03-06; 2023-12-14",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F23": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "title block",
   "quote": "Marc Marone Benjamin Van Durme Johns Hopkins University",
   "claim": "the authors' affiliation: Johns Hopkins University",
   "modality": "documented",
   "kind": "attribution",
   "kernel": false,
   "const": false,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F24": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§1 Introduction",
   "quote": "The critical property of a data portrait is membership inference: whether an example was part of a data collection.",
   "claim": "the critical property of a data portrait: membership inference, whether an example was part of a data collection",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": true,
   "lineage": "the membership question",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F25": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§1 Introduction",
   "quote": "In discussing these tools we consider three perspectives: content creators (owners), scientists, and content consumers. Creators wish to know whether a dataset contains their content (e.g. copywritten code).",
   "claim": "three perspectives: content creators, scientists, content consumers; creators want to know whether a dataset contains their content",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "uses: leakage, plagiarism, creators",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F26": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§2.1 Documentation Artifacts",
   "quote": "We view a Data Portrait as a complementary type of documentation artifact - one that answers the membership question.",
   "claim": "a Data Portrait as a documentation artifact complementing datasheets and model cards, answering the membership question",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "documentation artifacts: beside datasheets and model cards",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F27": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§3 Data Portraits",
   "quote": "We outline desirable properties of a dataset documentation tool built around membership inference, recognizing that some may induce tradeoffs. 1. Fast: An artifact should support low latency programmatic access. 2. Lightweight: An artifact should not be much more expensive than storing the original data. 3. Avoids Redistribution:",
   "claim": "the desirable properties of a membership tool: fast, lightweight, avoiding redistribution, local, indexing, match flexibility, with tradeoffs",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F28": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§3 Data Portraits (Other Implementations)",
   "quote": "Given these tradeoffs, we use a probabilistic approach for membership inference: a data sketch.",
   "claim": "after weighing grep, full-text indexes and suffix arrays, a probabilistic approach: a data sketch",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F29": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§3.1 Bloom Filter Sketching",
   "quote": "Bloom filters are similar to hash tables, but they store only the hash of a data element rather than storing the element itself. … Bloom filters can output false positives but never false negatives.",
   "claim": "a Bloom filter stores only hashes of elements; it can give false positives, never false negatives",
   "modality": "documented",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F30": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§3.2 Bloom Filter Construction",
   "quote": "This reduces space requirements but carries the risk of missing some string of interest if it is split by a tile boundary in the corpus.",
   "claim": "storing strided n-grams saves space and risks missing a string split by a tile boundary",
   "modality": "self-description",
   "kind": "limit",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F31": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§3.2 (Redistribution)",
   "quote": "Since Bloom filters distribute only hashes, they provide some obfuscation. … Note that in principle, proprietary LLM providers (e.g ChatGPT [29]) could release a Bloom filter based Data Portrait without revealing their exact training data – only the hashes of some substrings.",
   "claim": "distributing only hashes obfuscates the data; a proprietary provider could in principle release a portrait without revealing its training data",
   "modality": "interpretation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F32": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§4.1 Constructing Portraits",
   "quote": "Further examples in this paper use a sketch built on character n-grams of length 50. … In effect, a 50 character span is already a good document fingerprint.",
   "claim": "the sketch uses character n-grams of length 50, a span the authors find a good document fingerprint",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F33": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§4.1 The Pile",
   "quote": "The structure takes only 27 GB on disk and has a false positive rate of 1 × 10−3",
   "claim": "the Pile portrait: 27 GB on disk, false positive rate 1 × 10⁻³",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "size and speed",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F34": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§4.4 Code Generation",
   "quote": "This artifact exactly documents the training data used in StarCoder, a 15.5B parameter code generation model [25]. … This plugin implements a plagiarism check by calling our Portrait service.",
   "claim": "the Stack portrait documents StarCoder's training data exactly and powers a plugin's plagiarism check",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F35": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§4.4 Code Generation",
   "quote": "Our Portrait service acts as a lightweight first-pass check for model plagiarism and additional tools from Li et al. [25] can be used to further assess attribution.",
   "claim": "the service is a first-pass plagiarism check; attribution needs further tools",
   "modality": "self-description",
   "kind": "limit",
   "kernel": false,
   "const": false,
   "lineage": "uses: leakage, plagiarism, creators",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F36": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§4.2 WMT Overlap",
   "quote": "we cannot conclude whether translation targets are memorized by a specific model, but our Pile-based sketch shows that some test data has leaked.",
   "claim": "a finding and its limit: some WMT test data has leaked into the Pile; whether a specific model memorized it cannot be concluded",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "uses: leakage, plagiarism, creators",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F37": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§4.3 Overlap Detection",
   "quote": "Target summaries from XSum have the highest incidence of overlap with the Pile. … This suggests that some LDC files are publicly exposed and scraped into LLM corpora",
   "claim": "XSum summaries overlap the Pile most; some LDC files appear scraped into LLM corpora",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "uses: leakage, plagiarism, creators",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F38": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§5.1 False Positives",
   "quote": "We first find that there are no chained matches when querying our system with novel New York Times text. … we next find a false positive rate of 7 × 10−4",
   "claim": "on novel text, no chained matches and a measured false positive rate of 7 × 10⁻⁴",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "limits and error",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F39": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§6 Impact and Limitations",
   "quote": "Our work is limited in that we focus on studying only a few datasets. … We stress a lightweight and efficient hash-based approach to membership testing, but this comes with certain trade-offs such as false positives and the inability to retrieve the context surrounding a match.",
   "claim": "the stated limits: few datasets; false positives; no retrieval of the context around a match",
   "modality": "self-description",
   "kind": "limit",
   "kernel": true,
   "const": false,
   "lineage": "limits and error",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F40": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§6 Impact and Limitations",
   "quote": "That is, an open plagiarism detection tool could be used to avoid plagiarism detection.",
   "claim": "a stated risk: an open plagiarism detector could be used to evade detection",
   "modality": "self-description",
   "kind": "limit",
   "kernel": false,
   "const": false,
   "lineage": "limits and error",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F41": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§6 Impact and Limitations",
   "quote": "For example, a non-technical content creator could search for their material in a documented dataset.",
   "claim": "a hoped use: a non-technical content creator searching for their material in a documented dataset",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "uses: leakage, plagiarism, creators",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F42": {
   "source": "B10",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "§7 Conclusions and Future Work",
   "quote": "proposed the adoption of Data Portraits: records supporting membership inference on the data used to train a model. … call for dataset and model creators to release some form of Data Portrait as part of their essential documentation.",
   "claim": "the conclusion's definition: records supporting membership inference on a model's training data, and a call for their release as essential documentation",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "what it is: artifacts that record training data",
   "sense": "the entity",
   "date": "2023-12-14 (v2)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F43": {
   "source": "B1",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "page text",
   "quote": "For the Target space at TED 2017 in Vancouver, we designed and produced unique visual Data Portraits of the attendees: visual designs based on answers to simple but evocative questions that we turned into wearable buttons to spark conversations and connections for the five days of the event.",
   "claim": "homonym: Giorgia Lupi's (Accurat's) TED 2017 Data Portraits, wearable buttons designed from attendees' answers",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: Giorgia Lupi / Accurat, Data Portraits at TED 2017",
   "date": "2017 (the project); page undated",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F44": {
   "source": "B3",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "page text",
   "quote": "Book — Non-fiction. By the W. E. B. Du Bois Center at University of Massachusetts Amherst. 2018. 144 pages. … W. E . B. Du Bois’s Data Portraits is the first complete publication of W. E. B. Du Bois’s groundbreaking charts, graphs, and maps presented at the 1900 Paris Exposition.",
   "claim": "homonym: the 2018 book W. E. B. Du Bois's Data Portraits, the first complete publication of Du Bois's charts, graphs and maps of the 1900 Paris Exposition",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: W. E. B. Du Bois's data portraits (1900) and the 2018 book",
   "date": "2018 (the book); 1900 (the charts)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F45": {
   "source": "B4",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "page text",
   "quote": "For the Target-sponsored social space at TED 2017 in Vancouver, Giorgia Lupi designed and produced unique visual Data Portraits for attendees to spark conversations and connections among them during the five-day conference. … The team generated more than 500 data portraits during the five-day conference.",
   "claim": "homonym: Pentagram's account of Lupi's TED 2017 Data Portraits, more than 500 made",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: Giorgia Lupi, TED Data Portraits (Pentagram)",
   "date": "2017 (the project); page undated",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F46": {
   "source": "B5",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "card snippet (unreachable)",
   "quote": "Sociologist W.E.B. Du Bois created sixty hand-drawn \"data portraits\" between 1897 and 1990 with Atlanta University students and scholars, visualizing Black Amer...",
   "claim": "homonym: SparkMap's card on Du Bois's sixty hand-drawn data portraits with Atlanta University students (the snippet's '1990' as rendered)",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: W. E. B. Du Bois's data portraits (SparkMap)",
   "date": "undated (snippet only)",
   "quote_ok": true,
   "checked_against": "the card's title and snippet in the registry (cite_list), whitespace-normalised"
  },
  "F47": {
   "source": "B6",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "video card title (5m; University of Massachusetts Amherst Libraries)",
   "quote": "W. E. B. Du Bois's Data Portraits: Visualizing Black America",
   "claim": "homonym: a five-minute video by the UMass Amherst Libraries on the Du Bois book",
   "modality": "documented",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: W. E. B. Du Bois's Data Portraits (video)",
   "date": "undated (title only)",
   "quote_ok": true,
   "checked_against": "the card's title and snippet in the registry (cite_list), whitespace-normalised"
  },
  "F48": {
   "source": "B7",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "card snippet (unreachable)",
   "quote": "Data portraits depict their subjects' accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, soci...",
   "claim": "homonym: the MIT Media Lab's data portraits, depicting subjects' accumulated data in place of their faces",
   "modality": "documented",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: Judith Donath's data portraits (MIT Media Lab)",
   "date": "undated (snippet only)",
   "quote_ok": true,
   "checked_against": "the card's title and snippet in the registry (cite_list), whitespace-normalised"
  },
  "F49": {
   "source": "B8",
   "origin": "card (homonym)",
   "pool": "B_D",
   "locus": "article, ¶2; date",
   "quote": "June 30, 2014 … digital media scholar Judith Donath defines data portraits as “depictions of people made by visualizing data by and about them.”",
   "claim": "homonym: Judith Donath's definition of data portraits, as ARTnews reported it in 2014",
   "modality": "attributed",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the other entities of the name",
   "sense": "homonym: Judith Donath's data portraits (ARTnews)",
   "date": "2014-06-30",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F50": {
   "source": "B11",
   "origin": "wikipedia",
   "pool": "B_R",
   "locus": "index.php?title=Data_Portraits&action=raw (and Data_portraits, Data_portrait, the paper's title)",
   "quote": "Error: 404, Not Found",
   "claim": "the encyclopedia has no article titled Data Portraits, data portraits, data portrait, or the paper's title (HTTP 404 for each raw fetch)",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "the encyclopedia and the name",
   "sense": "the name (encyclopedia)",
   "date": "2026-11-01 (read)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11.txt (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F51": {
   "source": "B11",
   "origin": "wikipedia",
   "pool": "B_R",
   "locus": "api.php search for \"Data Portraits\"",
   "quote": "\"totalhits\":7",
   "claim": "the encyclopedia's search for the phrase returns seven articles (the Du Bois book cited in several; none on the paper's artifact)",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "the encyclopedia and the name",
   "sense": "the name (encyclopedia)",
   "date": "2026-11-01 (read)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-search.json (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  },
  "F52": {
   "source": "B12",
   "origin": "own_page",
   "pool": "B_R",
   "locus": "pile.dataportraits.org; stack.dataportraits.org; stack-v2.dataportraits.org",
   "quote": "error code: 502",
   "claim": "the three sketch front ends answered HTTP 502 on 2026-11-01",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "publication, lifecycle and access",
   "sense": "the entity",
   "date": "2026-11-01 (read)",
   "quote_ok": true,
   "checked_against": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-pile.html (whitespace-normalised; multi-part quotes joined by ' … ', each part checked)"
  }
 },
 "field_sources": {
  "B1": "giorgialupi — Data Portraits at TED — giorgialupi",
  "B2": "Data Portraits — Data Portraits",
  "B3": "Zinn Education Project — W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
  "B4": "Pentagram — TED Data Portraits - Pentagram",
  "B5": "SparkMap — W.E.B. Du Bois & Data Visualization History - SparkMap",
  "B6": "YouTube·University of Massachusetts Amherst Libraries — W. E. B. Du Bois's Data Portraits: Visualizing Black America",
  "B7": "MIT Media Lab — Data Portraits - MIT Media Lab",
  "B8": "ARTnews.com — 10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ...",
  "B9": "arXiv — [2303.03919] Data Portraits: Recording Foundation Model Training Data",
  "B10": "arXiv — Data Portraits: Recording Foundation Model Training Data (PDF, v2)",
  "B11": "Wikipedia — no article 'Data Portraits' (HTTP 404); search for \"Data Portraits\"",
  "B12": "dataportraits.org — the Pile, Stack and Stack V2 sketch front ends (HTTP 502)"
 },
 "delta": {
  "T_vs_LB": "At 'Data Portraits' (T1, eight cards, B1–B8, the disclosed pool B_D; B9–B12, the arXiv abstract, the paper, the encyclopedia's record of absence and the sketch front ends, are the reference pool and stand outside this comparison, §0.3), T composes a definition of the term, four senses under 'Key Examples and Meanings' and a closing offer. Substituted reading (collision): the composition's entity is the visualization term, defined as 'visual representations of people made by plotting their accumulated data' (T1-1) from card 7's snippet ('Data portraits depict their subjects' accumulated data rather than their faces', F48) and card 8's Donath definition (F49); the census's entity stands fourth (T1-5), from card 2 (B2), the one card of eight that concerns it. Of B_D's twelve claims on the entity (F1–F12, all on card 2's page, dataportraits.org), T composes one, F3, at a coarser grain: 'lightweight inspection tools … that record dataset training traces' for 'artifacts that record data and allow for inspection'. 'training traces' is the composition's phrase; 'lightweight' stands on no B_D source: card 2's page says 'minimal and efficient' (F6) and 'rapid, minimal' (F7), and 'lightweight' is the abstract's (F20) and the paper's, in B_R, which cannot show what the composition had (§0.3). Available, not composed: the question the artifact answers (F2), membership testing as the site's own description and its limit, 'nothing more' (F1, F6), data sketching (F5), the complement to documentation artifacts (F4), the front ends and their corpora, The Pile, The Stack and The Stack V2 (F7–F10), the offer to build on other data and the code (F11), and the two authors and the venue in the site's citation (F12). Composed without attribution, by grain: 'Contemporary research groups use the term' gives the artifact to unnamed groups where card 2's page names its two authors and its venue (F12); the recording function is composed at a grain coarser than its source's, so the grain rule records it and finds no breach. Limit dropped: 'nothing more' (F6). The homonym senses are composed from their cards at their grain or near it (delta.homonym_alignment): Du Bois from cards 3, 5 and 6, 'property ownership' on no saved card text; Donath from cards 7 and 8, 'subjective renderings' and 'travel patterns' on no saved card text (card 7 unreachable, its snippet truncated); Lupi from cards 1 and 4. Content replaced by use: the closing offer (T1-6), which names the artifact's sense 'AI data tracking tools'. Nothing is assigned to another; nothing is denied.",
  "T_vs_LB_reading": "v0.10 §0.5: the first diff is ontological. T's ontology at the name is the term's: a kind of portrait, its exemplars by person (Du Bois, Donath, Lupi), the artifact a fourth exemplar without a person. Of the one card that concerns the census's entity, it keeps the recording and drops what the card's own page makes central, the membership question and its 'nothing more', the corpora, the sketch and the makers' names.",
  "LB_vs_LBA": "Admission adds, in one claim space with the field's: the census's record questions, readback (TLL-12), name (TLL-11) and makes (TLL-13), its readings of the portrait as a corpus-side readback instrument that settles membership for a recorded corpus and is no proof of a model's use (TLL-20, TLL-21), its negative tags (TLL-26, TLL-27, TLL-28) and what it does not answer (TLL-19); #186's claim that the surveyed texts are in the corpora (O186-01), its persistence criterion (O186-02) and its partial measurability of reception (O186-03); the archive's model-side membership tests, the copyright trap, the membership attack and its divergence from a failed surfacing probe (O127-03, O127-04), its own canary phrase (O837-01) and distinguishability between models trained with and without a work (O1075-02); its accounts of the name in the training pipeline, head metadata discarded (O127-01), style unattributable (O127-02), the provenance envelope (O1075-01), semantic labour unattributed (O264-01), a corpus without provenance (O727-01) and provenance loss passed into the corpus (O939-01); its reading of datasheets and model cards (O623-01, O1075-03); #1543's question of what a text makes a machine able to write, the training loop and the separation of composition from provenance (O1543-01 to O1543-04); and the Capture Registry's findings at the name (CAP-1-01 to CAP-1-11). Not surfaced: no archive source is on T's rail (archive_controlled_cites 0). The field is carried whole; the archive adds to the field's own object the questions of the name and of what a model makes, which the field leaves to 'further tools' to assess attribution (F35) and to the context a hash cannot retrieve (F39).",
  "alignment": {
   "T1_cards_claims": [
    "F1",
    "F2",
    "F3",
    "F4",
    "F5",
    "F6",
    "F7",
    "F8",
    "F9",
    "F10",
    "F11",
    "F12",
    "F43",
    "F44",
    "F45",
    "F46",
    "F47",
    "F48",
    "F49"
   ],
   "T1_composed": [
    "F3 (at a coarser grain: 'lightweight inspection tools … that record dataset training traces')",
    "F43, F45 (Lupi's TED 2017 buttons)",
    "F44, F46, F47 (Du Bois's charts for the 1900 Paris Exposition)",
    "F48, F49 (Donath's data portraits; the definition T1-1)"
   ],
   "T1_composed_note": "of the entity's twelve B_D claims, F3 alone, at a coarser grain; the homonyms' kernels composed at their grain or near it",
   "T1_available_not_composed": [
    "F1",
    "F2",
    "F4",
    "F5",
    "F6",
    "F7",
    "F8",
    "F9",
    "F10",
    "F11",
    "F12"
   ],
   "T1_outside_cards": [
    "T1-5 'lightweight': on no B_D source (F20, the abstract's, is in B_R)",
    "T1-5 'training traces': on no source",
    "T1-5 'Contemporary research groups': on no source; F12 names two authors",
    "T1-2 'property ownership': on no saved card text",
    "T1-3 'subjective renderings', 'travel patterns', 'social networks': on no saved card text (B7 unreachable; its snippet is cut at 'soci...')",
    "T1-6 the closer: no source"
   ],
   "field_not_surfaced": [
    "F13",
    "F14",
    "F15",
    "F16",
    "F17",
    "F18",
    "F19",
    "F20",
    "F21",
    "F22",
    "F23",
    "F24",
    "F25",
    "F26",
    "F27",
    "F28",
    "F29",
    "F30",
    "F31",
    "F32",
    "F33",
    "F34",
    "F35",
    "F36",
    "F37",
    "F38",
    "F39",
    "F40",
    "F41",
    "F42",
    "F50",
    "F51",
    "F52"
   ],
   "rule": "a field claim counts as composed when T states it at its source's grain or coarser; B_D only (B9–B12 excluded, §0.3); the transcript read for this alignment only"
  },
  "collision": {
   "reached": "collision (per_dim.reached)",
   "composition_entity": "the visualization term, with the entity as the fourth of four senses",
   "other_entities_composed": [
    "W. E. B. Du Bois's data portraits for the 1900 Paris Exposition and the 2018 book (T1-2; B3, B5, B6)",
    "Judith Donath's data portraits (T1-1, T1-3; B7, B8)",
    "Giorgia Lupi's Data Portraits at TED 2017 (T1-4; B1, B4)"
   ],
   "basis": [
    "CAP-1-01",
    "CAP-1-03",
    "CAP-1-05",
    "CAP-1-06",
    "CAP-1-10"
   ]
  },
  "homonym_alignment": [
   {
    "T": "T1-1",
    "cards": [
     "B7",
     "B8"
    ],
    "spans": [
     {
      "source": "B7",
      "quote": "Data portraits depict their subjects' accumulated data rather than their faces.",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     },
     {
      "source": "B8",
      "quote": "depictions of people made by visualizing data by and about them",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     }
    ],
    "note": "the term's definition at the cards' grain; 'plotting' and 'drawing or photographing their physical faces' the composition's"
   },
   {
    "T": "T1-2",
    "cards": [
     "B3",
     "B5",
     "B6"
    ],
    "spans": [
     {
      "source": "B3",
      "quote": "charts, graphs, and maps presented at the 1900 Paris Exposition",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     },
     {
      "source": "B5",
      "quote": "sixty hand-drawn \"data portraits\"",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     }
    ],
    "note": "'over 60' for the snippet's 'sixty'; 'education' from B3 ('advances in education'); 'property ownership' on no saved card text"
   },
   {
    "T": "T1-3",
    "cards": [
     "B7",
     "B8"
    ],
    "spans": [
     {
      "source": "B8",
      "quote": "digital media scholar Judith Donath defines data portraits",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     },
     {
      "source": "B7",
      "quote": "visualizations of discussion contributions, browsing histories",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     }
    ],
    "note": "'subjective renderings', 'social networks' and 'travel patterns' on no saved card text; B7 unreachable, its snippet cut at 'soci...'"
   },
   {
    "T": "T1-4",
    "cards": [
     "B1",
     "B4"
    ],
    "spans": [
     {
      "source": "B1",
      "quote": "turned into wearable buttons",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     },
     {
      "source": "B4",
      "quote": "Responses to simple but evocative questions provided the input to generate the data portraits.",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     }
    ],
    "note": "'visual languages' after B1's and B4's 'the visual language of the Data Portraits'"
   },
   {
    "T": "T1-5",
    "cards": [
     "B2"
    ],
    "spans": [
     {
      "source": "B2",
      "quote": "artifacts that record data and allow for inspection",
      "quote_ok": true,
      "checked_against": "the saved text or the card snippet"
     }
    ],
    "note": "the entity, at a coarser grain; 'lightweight' and 'training traces' on no B_D source"
   },
   {
    "T": "T1-6",
    "cards": [],
    "spans": [],
    "note": "the closer, content replaced by use (N_ext)"
   }
  ],
  "pool_note": "T against L_F(B_D): B1–B8; L_F(B) against L(B ∪ A): B1–B12 with A"
 },
 "kernel": [
  {
   "K": "K1",
   "claim": "`TLL-11`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "interpretation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: no field claim says whether a portrait keeps a recorded work's name; F29 (hashes only) and F39 (no context around a match) bear on it",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K2",
   "claim": "`TLL-13`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "interpretation",
   "sense": "what a model makes",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: the field states leakage and plagiarism checks (F19, F35) and no claim on what a model makes from a recorded work",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K3",
   "claim": "`TLL-19`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "interpretation",
   "sense": "limits",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F36 (the WMT finding stops short of a specific model) and F35 (attribution needs further tools)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K4",
   "claim": "`TLL-20`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "interpretation",
   "sense": "membership and the training layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F24 (membership inference as the critical property), read as settling corpus membership",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K5",
   "claim": "`TLL-21`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "interpretation",
   "sense": "limits",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F36 (leakage shown, a specific model's memorization not concluded)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K6",
   "claim": "`TLL-26`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "stipulation",
   "sense": "the corpora it records",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none: the field states the corpora recorded (F18) and no claim on the artifact as a corpus",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K7",
   "claim": "`TLL-27`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "stipulation",
   "sense": "limits",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F2, F14 (the question is put of the model, answered from the recorded data)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K8",
   "claim": "`TLL-28`",
   "source": "traininglayerliterature.org layer census record",
   "M_src": "stipulation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: F29 (hashes only); no field claim on attribution metadata",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K9",
   "claim": "`O127-01`",
   "source": "#127 Inscriptions That Survive the Tokenizer: SPXI-TLP v2.2 — The Training-Layer Prov",
   "M_src": "interpretation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: no field claim on what a corpus keeps of a page's metadata",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K10",
   "claim": "`O127-02`",
   "source": "#127 Inscriptions That Survive the Tokenizer: SPXI-TLP v2.2 — The Training-Layer Prov",
   "M_src": "interpretation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K11",
   "claim": "`O127-03`",
   "source": "#127 Inscriptions That Survive the Tokenizer: SPXI-TLP v2.2 — The Training-Layer Prov",
   "M_src": "attributed",
   "sense": "membership and the training layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F24 (membership inference on a data collection, corpus side) beside a model-side attack",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K12",
   "claim": "`O127-04`",
   "source": "#127 Inscriptions That Survive the Tokenizer: SPXI-TLP v2.2 — The Training-Layer Prov",
   "M_src": "stipulation",
   "sense": "membership and the training layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K13",
   "claim": "`O186-01`",
   "source": "#186 Who Is Writing for the Machines? Training-Layer Literature: A Field Survey",
   "M_src": "interpretation",
   "sense": "membership and the training layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F24, F41 (a creator could search for their material in a documented dataset)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K14",
   "claim": "`O186-02`",
   "source": "#186 Who Is Writing for the Machines? Training-Layer Literature: A Field Survey",
   "M_src": "stipulation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: no field claim on metadata that lets a work be found and attributed",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K15",
   "claim": "`O186-03`",
   "source": "#186 Who Is Writing for the Machines? Training-Layer Literature: A Field Survey",
   "M_src": "interpretation",
   "sense": "what a model makes",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K16",
   "claim": "`O264-01`",
   "source": "#264 Semantic Exhaustion: An Executive Summary — The Depletion Threshold for Meaning-",
   "M_src": "interpretation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K17",
   "claim": "`O623-01`",
   "source": "#623 Semantic Infrastructure: From Tim Berners-Lee to the Semantic Economy",
   "M_src": "interpretation",
   "sense": "a documentation artifact",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F4, F26 (a portrait as a documentation artifact beside datasheets)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K18",
   "claim": "`O727-01`",
   "source": "#727 Political Economy Has Always Already Been Semantic Economy: Metadata Packet for ",
   "M_src": "interpretation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K19",
   "claim": "`O837-01`",
   "source": "#837 The Josephus Thesis Is Not the Jesus Myth Thesis: Preemptive Disambiguation MPAI",
   "M_src": "stipulation",
   "sense": "membership and the training layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F24 (membership on the corpus side) beside a model-side canary",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K20",
   "claim": "`O939-01`",
   "source": "#939 EA-PROVENANCE-DEBT-01 v0.2: Provenance Debt and the Extraction Economy of Unmark",
   "M_src": "interpretation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K21",
   "claim": "`O1075-01`",
   "source": "#1075 EA-NEGSHAPE-01 v0.2: The Negative Shape of the Work",
   "M_src": "stipulation",
   "sense": "the name of a recorded work",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: F29 (hashes of spans) records text units and no metadata unit",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K22",
   "claim": "`O1075-02`",
   "source": "#1075 EA-NEGSHAPE-01 v0.2: The Negative Shape of the Work",
   "M_src": "attributed",
   "sense": "membership and the training layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F36 (corpus-side leakage, model-side memorization not concluded)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K23",
   "claim": "`O1075-03`",
   "source": "#1075 EA-NEGSHAPE-01 v0.2: The Negative Shape of the Work",
   "M_src": "interpretation",
   "sense": "a documentation artifact",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F4, F26",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K24",
   "claim": "`O1543-01`",
   "source": "#1543 Generative Uptake: Training-Layer Literature Extended from Machine Reading to Ma",
   "M_src": "stipulation",
   "sense": "what a model makes",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "missing distinction: the field has no claim on what a model writes after encountering a recorded work",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K25",
   "claim": "`O1543-02`",
   "source": "#1543 Generative Uptake: Training-Layer Literature Extended from Machine Reading to Ma",
   "M_src": "stipulation",
   "sense": "what a model makes",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K26",
   "claim": "`O1543-03`",
   "source": "#1543 Generative Uptake: Training-Layer Literature Extended from Machine Reading to Ma",
   "M_src": "stipulation",
   "sense": "what a model makes",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "qualified claim: F35 (plagiarism first pass; attribution needs further tools)",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  },
  {
   "K": "K27",
   "claim": "`O1543-04`",
   "source": "#1543 Generative Uptake: Training-Layer Literature Extended from Machine Reading to Ma",
   "M_src": "stipulation",
   "sense": "what a model makes",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not (uptake §8.3))",
   "contrast": "none",
   "resolver": "a later capture at 'Data Portraits' (ADDR-1ac4e7c48a77) on Google AI Overview, seated in the Capture Registry (§8.3)"
  }
 ],
 "kernel_T": [
  {
   "claim": "F2",
   "note": "the question ('has the model already encountered a given example during training?'), card 2's lead, absent"
  },
  {
   "claim": "F6",
   "note": "'membership testing and nothing more', the site's own limit, absent"
  },
  {
   "claim": "F8",
   "note": "The Pile sketch, absent"
  },
  {
   "claim": "F9",
   "note": "The Stack sketch, absent"
  },
  {
   "claim": "F10",
   "note": "The Stack V2 sketch, absent"
  },
  {
   "claim": "F12",
   "note": "the authors and venue in the site's citation, absent; T says 'Contemporary research groups'"
  },
  {
   "claim": "F1",
   "note": "'Dataset documentation through membership testing', absent"
  },
  {
   "claim": "F5",
   "note": "data sketching, absent"
  },
  {
   "claim": "F4",
   "note": "the complement to documentation artifacts, absent"
  },
  {
   "claim": "F11",
   "note": "build on your data; code on GitHub, absent"
  }
 ],
 "ledger": {
  "archive": "datasets/negative-of-the-negative/v2/ledgers/data-portraits/ledger-archive.json",
  "field": "datasets/negative-of-the-negative/v2/ledgers/data-portraits/field.json",
  "selection": "datasets/negative-of-the-negative/v2/traversal/data-portraits/",
  "config": "datasets/negative-of-the-negative/v2/panel/configs/data-portraits.json",
  "intake": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/",
  "audit": null
 },
 "ontology": {
  "composition": "equal",
  "composition_note": "equal-terms case (v0.9 §0.9, §4.1, unchanged in v0.10): one claim space; the field's and the archive's claims ordered together by genealogy within each sense; the field's claims composed as received, the archive's as claims on the same object",
  "case": {
   "value": "equal",
   "statement": "The archive reads the field's entity as the entity. The census record describes Data Portraits from the paper's abstract and keeps the resource's terms (operator, kind, unit, record object, method) apart from its own reading (tll_reading), which places the portrait on the corpus side of #186's membership question and leaves the name and #1543's question of what a model makes outside it. The archive's deposits that bear on the classes the field and the census source for the entity (training-data membership testing, the training corpora, documentation artifacts) add model-side membership tests, the archive's canary, the provenance envelope, the loss of head metadata before training, and the reading of datasheets as semantic accounting. None of these reads Data Portraits as made by reception or contests that it is the paper's membership-testing artifact: they add senses (readback, name, makes) on the field's object. ρ = extends; the equal-terms case.",
   "loci": [
    {
     "dep": "TLL",
     "locus": "census.json rows[slug == 'data-portraits'].tll_reading",
     "quote": "A readback instrument held on the corpus side: a portrait records the dataset so that a work can be tested for membership, which settles #186's 'is the work in the layer' for a recorded corpus."
    },
    {
     "dep": "TLL",
     "locus": "census.json rows[slug == 'data-portraits'].tll_reading",
     "quote": "Dataset membership is not proof that a particular model used a passage, and the name and what a model made from the work (#1543's question) are outside what the saved abstract records."
    },
    {
     "dep": 186,
     "locus": "Who is in the field? ¶ (line 54)",
     "quote": "Each of these practices touches the training layer — every one of these texts is, as a matter of fact, in the corpora."
    },
    {
     "dep": 1543,
     "locus": "§1 The claim (line 81)",
     "quote": "The test is no longer only \"can a machine correctly identify what this text means?\" but \"what does this text make possible for a machine to write?\""
    },
    {
     "dep": 127,
     "locus": "§IV.2 HTML metadata (line 463)",
     "quote": "HTML head metadata — author tags, copyright, Open Graph, Twitter cards — is parsed by search engines and social media unfurl bots, but is almost always discarded by readability extractors that prepare text for training corpus inclusion. The training-data pipeline retains the *body text* and the *URL* (sometimes)."
    }
   ]
  },
  "E_F": {
   "type": "C (public entity: a dataset-documentation artifact and membership-testing tool from a research paper)",
   "definition": "Data Portraits, artifacts that record the data a foundation model is trained on and allow downstream inspection by membership testing, proposed in a 2023 paper by Marc Marone and Benjamin Van Durme (Johns Hopkins University; NeurIPS 2023 Datasets and Benchmarks), implemented as Bloom-filter sketches of hashed, strided 50-character n-grams (about 3% of a dataset's size; false positives, never false negatives), built for The Pile and The Stack (StarCoder's training data exactly) with a sketch of The Stack V2 listed at dataportraits.org; it answers test-set leakage and plagiarism questions, and its makers state its limits (few datasets, false positives, no context around a match, evasion); the name is shared with Du Bois's, Donath's and Lupi's data portraits, and the encyclopedia has no article on it",
   "claims": [
    "F3",
    "F15",
    "F13",
    "F23",
    "F22",
    "F24",
    "F6",
    "F29",
    "F32",
    "F20",
    "F18",
    "F34",
    "F10",
    "F19",
    "F39",
    "F44",
    "F49",
    "F43",
    "F50"
   ]
  },
  "E_A": {
   "definition": "the same artifact, in the traininglayerliterature.org layer census record and the archive's deposits: a readback instrument held on the corpus side, a portrait recording a dataset so that a work can be tested for membership, which settles for a recorded corpus whether a work is in the training layer and is no proof that a particular model used a passage; no corpus, no model-side test and no record of authorship, the name of a recorded work and what a model made from it lying outside its record; set beside the archive's model-side membership tests (the copyright trap, the membership attack and its divergence from a failed surfacing probe, the archive's canary), its account of the name lost before training (head metadata discarded, the provenance envelope, provenance loss passed into the corpus) and #1543's question of what a text makes a machine able to write",
   "claims": [
    "TLL-20",
    "TLL-21",
    "TLL-11",
    "TLL-12",
    "TLL-13",
    "TLL-26",
    "TLL-27",
    "TLL-28",
    "O186-01",
    "O127-01",
    "O127-03",
    "O127-04",
    "O837-01",
    "O1075-01",
    "O1543-01",
    "O1543-03"
   ],
   "defining_papers": [
    186,
    1543,
    127,
    1075,
    837,
    623,
    939,
    727,
    264
   ],
   "defining_records": [
    "traininglayerliterature.org layer census record (census.json rows[slug == 'data-portraits'], organizing_question, admission_rule, areas, anchor), https://www.traininglayerliterature.org/aggregators/data-portraits",
    "Capture Registry entry data-portraits-aio-20261101 (OBS-103202920b0d), https://www.alexanarch.org/captures/data-portraits-aio-20261101/"
   ]
  },
  "rho": {
   "value": "extends",
   "status": "read from the census record, the capture record and the deposits that bear on the entity's classes (#186 and #1543, the census anchors; #127, #1075, #837, #623, #939, #727, #264), 2026-11-01",
   "statement": "The archive reads Data Portraits as its own pages and its paper present it and adds, on the same object, the census's three record questions: readback, a membership query that settles corpus membership for a recorded corpus (TLL-12, TLL-20); the name, not recorded (TLL-11, TLL-28), beside the archive's account of metadata discarded before training and of provenance in the corpus (O127-01, O127-02, O1075-01, O264-01, O727-01, O939-01, O186-02); and what a model makes, outside the portrait (TLL-13, TLL-21), with #1543's question (O1543-01 to O1543-04). It sets the corpus-side test beside its model-side tests (O127-03, O127-04, O837-01, O1075-02) and reads datasheets and model cards as documentation norms and semantic accounting (O1075-03, O623-01). No archive claim contradicts a field claim.",
   "claims": [
    "TLL-12",
    "TLL-20",
    "TLL-11",
    "TLL-28",
    "TLL-13",
    "TLL-21",
    "O127-01",
    "O1075-01",
    "O1543-01",
    "O127-03",
    "O837-01",
    "O623-01"
   ],
   "guard": "ρ = extends claims an addition and no more: the archive's claims do not make the field's sources wrong at their dates; where the census's saved source was the abstract alone (TLL-11, TLL-17), the paper's fuller statements (F29, F39) stand beside it with their dates, and the name question stays a question (open block).",
   "considered": [
    {
     "value": "constructs",
     "rejected": "the archive does not read Data Portraits as made by reception; its record takes the paper's account of the artifact as the account"
    },
    {
     "value": "revises",
     "rejected": "Data Portraits is not the archive's own entity; no archive text restates it"
    }
   ]
  },
  "translation": [
   {
    "lineage": "what it is: artifacts that record training data",
    "field_claims": [
     "F3",
     "F15",
     "F13",
     "F23",
     "F12",
     "F1",
     "F42"
    ],
    "received": "F3 (B2, lead ¶2): \"We propose a widespread adoption of Data Portraits: artifacts that record data and allow for inspection.\" | F15 (B9, Abstract): \"We therefore propose a widespread adoption of Data Portraits: artifacts that record training data and allow for downstream inspection.\" | F13 (B9, title; authors): \"Title: Data Portraits: Recording Foundation Model Training Data … Authors: Marc Marone , Benjamin Van Durme\" | F23 (B10, title block): \"Marc Marone Benjamin Van Durme Johns Hopkins University\" | F12 (B2, Citing (BibTeX)): \"author = {Marone, Marc and {Van Durme}, Benjamin}, booktitle={Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track},\" | F1 (B2, <meta name=\"description\"> (the page head)): \"Dataset documentation through membership testing.\" | F42 (B10, §7 Conclusions and Future Work): \"proposed the adoption of Data Portraits: records supporting membership inference on the data used to train a model. … call for dataset and model creators to release some form of Data Portrait as part of their essential documentation.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-01, TLL-02, TLL-04, TLL-05, TLL-23",
    "archive_claims": [
     "TLL-01",
     "TLL-02",
     "TLL-04",
     "TLL-05",
     "TLL-23"
    ]
   },
   {
    "lineage": "corpora recorded: The Pile, The Stack, The Stack V2",
    "field_claims": [
     "F7",
     "F18",
     "F8",
     "F9",
     "F34",
     "F10",
     "F11"
    ],
    "received": "F7 (B2, Our Implementations): \"We host front-ends over several data sketches. These allow for rapid, minimal, membership testing over popular web-based text datasets.\" | F18 (B9, Abstract): \"Using our tools, we document a popular language modeling corpus (The Pile) and a recently released code modeling dataset (The Stack).\" | F8 (B2, Our Implementations: The Pile): \"Check if some text appears in a large language modeling corpus. This is a 50 character n-gram sketch on the Pile.\" | F9 (B2, Our Implementations: The Stack): \"Check if a code snippet appears in The Stack, a code dataset by Big Code.\" | F34 (B10, §4.4 Code Generation): \"This artifact exactly documents the training data used in StarCoder, a 15.5B parameter code generation model [25]. … This plugin implements a plagiarism check by calling our Portrait service.\" | F10 (B2, Our Implementations: The Stack V2): \"Check if text appears in The Stack V2. This dataset contains code and text about code, like documentation and papers.\" | F11 (B2, Our Implementations: Your Dataset): \"Code on Github! … if you'd like us to build our system on your data.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-09, TLL-10, TLL-30, TLL-26",
    "archive_claims": [
     "TLL-09",
     "TLL-10",
     "TLL-30",
     "TLL-26"
    ]
   },
   {
    "lineage": "the membership question",
    "field_claims": [
     "F2",
     "F14",
     "F24",
     "F6"
    ],
    "received": "F2 (B2, lead ¶1): \"Foundation models are trained on increasingly immense and opaque datasets. … it can be difficult to answer the straight forward question: has the model already encountered a given example during training?\" | F14 (B9, Abstract): \"Foundation models are trained on increasingly immense and opaque datasets. Even while these models are now key in AI system building, it can be difficult to answer the straightforward question: has the model already encountered a given example during training?\" | F24 (B10, §1 Introduction): \"The critical property of a data portrait is membership inference: whether an example was part of a data collection.\" | F6 (B2, lead ¶2): \"Our implementation is minimal and efficient in that it supports membership testing and nothing more.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-24, TLL-08, TLL-12, TLL-20, O186-01",
    "archive_claims": [
     "TLL-24",
     "TLL-08",
     "TLL-12",
     "TLL-20",
     "O186-01"
    ]
   },
   {
    "lineage": "documentation artifacts: beside datasheets and model cards",
    "field_claims": [
     "F21",
     "F4",
     "F26"
    ],
    "received": "F21 (B9, Abstract): \"We release a live interface of our tools at this https URL and call on dataset and model creators to release Data Portraits as a complement to current documentation practices.\" | F4 (B2, lead ¶2): \"These are a complement to existing forms of data and model documentation artifacts.\" | F26 (B10, §2.1 Documentation Artifacts): \"We view a Data Portrait as a complementary type of documentation artifact - one that answers the membership question.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside O1075-03, O623-01",
    "archive_claims": [
     "O1075-03",
     "O623-01"
    ]
   },
   {
    "lineage": "the sketch: Bloom filter, strided n-grams, hashes",
    "field_claims": [
     "F5",
     "F16",
     "F27",
     "F28",
     "F17",
     "F29",
     "F32",
     "F30",
     "F31"
    ],
    "received": "F5 (B2, lead ¶2): \"We introduce our solution based on data sketching (compressed and approximate views of large data).\" | F16 (B9, Abstract): \"First we outline the properties of such an artifact and discuss how existing solutions can be used to increase transparency.\" | F27 (B10, §3 Data Portraits): \"We outline desirable properties of a dataset documentation tool built around membership inference, recognizing that some may induce tradeoffs. 1. Fast: An artifact should support low latency programmatic access. 2. Lightweight: An artifact should not be much more expensive than storing the original data. 3. Avoids Redistribution:\" | F28 (B10, §3 Data Portraits (Other Implementations)): \"Given these tradeoffs, we use a probabilistic approach for membership inference: a data sketch.\" | F17 (B9, Abstract): \"We then propose and implement a solution based on data sketching, stressing fast and space efficient querying.\" | F29 (B10, §3.1 Bloom Filter Sketching): \"Bloom filters are similar to hash tables, but they store only the hash of a data element rather than storing the element itself. … Bloom filters can output false positives but never false negatives.\" | F32 (B10, §4.1 Constructing Portraits): \"Further examples in this paper use a sketch built on character n-grams of length 50. … In effect, a 50 character span is already a good document fingerprint.\" | F30 (B10, §3.2 Bloom Filter Construction): \"This reduces space requirements but carries the risk of missing some string of interest if it is split by a tile boundary in the corpus.\" | F31 (B10, §3.2 (Redistribution)): \"Since Bloom filters distribute only hashes, they provide some obfuscation. … Note that in principle, proprietary LLM providers (e.g ChatGPT [29]) could release a Bloom filter based Data Portrait without revealing their exact training data – only the hashes of some substrings.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-14",
    "archive_claims": [
     "TLL-14"
    ]
   },
   {
    "lineage": "size and speed",
    "field_claims": [
     "F33",
     "F20"
    ],
    "received": "F33 (B10, §4.1 The Pile): \"The structure takes only 27 GB on disk and has a false positive rate of 1 × 10−3\" | F20 (B9, Abstract): \"Our tool is lightweight and fast, costing only 3% of the dataset size in overhead.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-15",
    "archive_claims": [
     "TLL-15"
    ]
   },
   {
    "lineage": "uses: leakage, plagiarism, creators",
    "field_claims": [
     "F19",
     "F36",
     "F37",
     "F35",
     "F25",
     "F41"
    ],
    "received": "F19 (B9, Abstract): \"We show that our solution enables answering questions about test set leakage and model plagiarism.\" | F36 (B10, §4.2 WMT Overlap): \"we cannot conclude whether translation targets are memorized by a specific model, but our Pile-based sketch shows that some test data has leaked.\" | F37 (B10, §4.3 Overlap Detection): \"Target summaries from XSum have the highest incidence of overlap with the Pile. … This suggests that some LDC files are publicly exposed and scraped into LLM corpora\" | F35 (B10, §4.4 Code Generation): \"Our Portrait service acts as a lightweight first-pass check for model plagiarism and additional tools from Li et al. [25] can be used to further assess attribution.\" | F25 (B10, §1 Introduction): \"In discussing these tools we consider three perspectives: content creators (owners), scientists, and content consumers. Creators wish to know whether a dataset contains their content (e.g. copywritten code).\" | F41 (B10, §6 Impact and Limitations): \"For example, a non-technical content creator could search for their material in a documented dataset.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-18",
    "archive_claims": [
     "TLL-18"
    ]
   },
   {
    "lineage": "limits and error",
    "field_claims": [
     "F38",
     "F39",
     "F40"
    ],
    "received": "F38 (B10, §5.1 False Positives): \"We first find that there are no chained matches when querying our system with novel New York Times text. … we next find a false positive rate of 7 × 10−4\" | F39 (B10, §6 Impact and Limitations): \"Our work is limited in that we focus on studying only a few datasets. … We stress a lightweight and efficient hash-based approach to membership testing, but this comes with certain trade-offs such as false positives and the inability to retrieve the context surrounding a match.\" | F40 (B10, §6 Impact and Limitations): \"That is, an open plagiarism detection tool could be used to avoid plagiarism detection.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-17, TLL-19, TLL-27, TLL-21",
    "archive_claims": [
     "TLL-17",
     "TLL-19",
     "TLL-27",
     "TLL-21"
    ]
   },
   {
    "lineage": "publication, lifecycle and access",
    "field_claims": [
     "F22",
     "F52"
    ],
    "received": "F22 (B9, Comments; Submission history): \"NeurIPS 2023 Datasets and Benchmarks … [v1] Mon, 6 Mar 2023 04:22:33 UTC (651 KB) … [v2] Thu, 14 Dec 2023 16:55:42 UTC (614 KB)\" | F52 (B12, pile.dataportraits.org; stack.dataportraits.org; stack-v2.dataportraits.org): \"error code: 502\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside TLL-25, TLL-16, TLL-31, TLL-03",
    "archive_claims": [
     "TLL-25",
     "TLL-16",
     "TLL-31",
     "TLL-03"
    ]
   },
   {
    "lineage": "the other entities of the name",
    "field_claims": [
     "F44",
     "F46",
     "F47",
     "F49",
     "F48",
     "F43",
     "F45"
    ],
    "received": "F44 (B3, page text): \"Book — Non-fiction. By the W. E. B. Du Bois Center at University of Massachusetts Amherst. 2018. 144 pages. … W. E . B. Du Bois’s Data Portraits is the first complete publication of W. E. B. Du Bois’s groundbreaking charts, graphs, and maps presented at the 1900 Paris Exposition.\" | F46 (B5, card snippet (unreachable)): \"Sociologist W.E.B. Du Bois created sixty hand-drawn \"data portraits\" between 1897 and 1990 with Atlanta University students and scholars, visualizing Black Amer...\" | F47 (B6, video card title (5m; University of Massachusetts Amherst Libraries)): \"W. E. B. Du Bois's Data Portraits: Visualizing Black America\" | F49 (B8, article, ¶2; date): \"June 30, 2014 … digital media scholar Judith Donath defines data portraits as “depictions of people made by visualizing data by and about them.”\" | F48 (B7, card snippet (unreachable)): \"Data portraits depict their subjects' accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, soci...\" | F43 (B1, page text): \"For the Target space at TED 2017 in Vancouver, we designed and produced unique visual Data Portraits of the attendees: visual designs based on answers to simple but evocative questions that we turned into wearable buttons to spark conversations and connections for the five days of the event.\" | F45 (B4, page text): \"For the Target-sponsored social space at TED 2017 in Vancouver, Giorgia Lupi designed and produced unique visual Data Portraits for attendees to spark conversations and connections among them during the five-day conference. … The team generated more than 500 data portraits during the five-day conference.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received",
    "archive_claims": []
   },
   {
    "lineage": "the encyclopedia and the name",
    "field_claims": [
     "F50",
     "F51"
    ],
    "received": "F50 (B11, index.php?title=Data_Portraits&action=raw (and Data_portraits, Data_portrait, the paper's title)): \"Error: 404, Not Found\" | F51 (B11, api.php search for \"Data Portraits\"): \"\"totalhits\":7\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received",
    "archive_claims": []
   }
  ],
  "receiving_texts": "equal-terms case: every field claim composed as received, beside the archive claims it stands with; no receiving text"
 },
 "measure": {
  "rule": "v0.9 §4.10 (unchanged in v0.10): KO_B sentences surviving in the KO at similarity ≥ 0.95, with positions (computed as apply_v09.py does: difflib.SequenceMatcher ratio, the best KO sentence for each KO_B sentence)",
  "v0.10": {
   "surviving": 0,
   "of": 11,
   "positions": [
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null
   ]
  },
  "v0.9": {
   "surviving": 0,
   "of": 11,
   "positions": [
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null
   ]
  },
  "opening_run": {
   "rule": "§5.10 (equal-terms case): the field-only KO's sentences as the KO's opening run in their own order",
   "breach": false
  },
  "blocks": {
   "rule": "§4.1 (equal-terms case): the KO does not open with the field as a block or close with the archive as one; sentences typed by the ledgers of the claims they cite (F field-only, A archive-only, M both)",
   "sequence": "MMMFFAMFFMMAFFFAFFAAAAAMAAAAAAMFAAFFAAMM",
   "field_only_head": 0,
   "archive_only_tail": 0,
   "breach": false
  },
  "max_similarity": [
   0.339,
   0.407,
   0.522,
   0.18,
   0.6,
   0.546,
   0.289,
   0.634,
   0.657,
   0.633,
   0.208
  ],
  "ko_sentence_1_max_similarity_to_KO_B": 0.339
 },
 "homonyms": {
  "names": [
   "W. E. B. Du Bois's data portraits: the charts, graphs and maps of the 1900 Paris Exposition, and the 2018 book W. E. B. Du Bois's Data Portraits: Visualizing Black America (Princeton Architectural Press) (T1-2; B3, B5, B6)",
   "Judith Donath's data portraits, 'depictions of people made by visualizing data by and about them' (MIT Media Lab; The Social Machine) (T1-1, T1-3; B7, B8)",
   "Giorgia Lupi's Data Portraits at TED 2017 (Accurat; Pentagram), wearable buttons (T1-4; B1, B4)"
  ],
  "basis": "A collision at the name (per_dim.reached 'collision'): the composition defines the term as a kind of portrait and lists four senses, the artifact last (CAP-1-01, CAP-1-10); three other entities from seven of the eight cards. Each stays a distinct entity in the disambiguation, never a sense of the artifact's name; the encyclopedia has no article under the name (F50) and its search finds the Du Bois book (F51).",
  "distinct_entities": [
   "Marc Marone and Benjamin Van Durme, the authors (B2, B9, B10), persons",
   "Johns Hopkins University, their affiliation (B10)",
   "The Pile (EleutherAI) and The Stack, The Stack V2 (BigCode), the corpora recorded (B2, B9, B10); The Pile a row of its own in the census (the-pile)",
   "StarCoder, the model whose training data the Stack portrait documents (B10)",
   "OLMoTrace, the census's neighbour (TLL-29), a row of its own (datasets/negative-of-the-negative/v2/rows/olmotrace.json)",
   "datasheets for datasets and model cards, the documentation artifacts it complements (B10)",
   "The Stack, Benjamin Bratton's book, a homonym of the dataset in the archive's texts"
  ]
 },
 "notes": [
  "One address, Google AI Overview (expanded), signed out, incognito, 2026-11-01: data-portraits-aio-20261101 (OBS-103202920b0d, ADDR-1ac4e7c48a77, AXN:0BAD.CAPTURE.🌇🕘📎☿🔽🚀), eight cards, 158 words in the answer body with a closing offer, 10 markers. A NEW address (CAP-1-09); reached 'collision' (per_dim): the artifact the last of four senses of the name.",
  "Fetching: every reachable card saved verbatim with curl and the census agent 'Mozilla/5.0 (compatible; traininglayerliterature-census)' on 2026-11-01 between 15:13 and 15:15 UTC, raw and text with sha256 in field.json, the log in tries/fetch.log. Card 2 (dataportraits.org) was the entity's own site, unsaved by the census. The MIT Media Lab card (B7) failed TLS verification at both its URLs (self-signed certificate in chain), one retry with the proxy's CA bundle named, verification never disabled; SparkMap (B5) found no URL; the video card (B6) admitted by title. The paper's PDF (B10) and the three sketch front ends (B12, HTTP 502) were fetched as the entity's own pages, the census's unchecked items.",
  "Pools (v0.10 §0.3, §3.1(a)): B_D = B1–B8 (19 claims: F1–F12 on the entity, from card 2's page, and F43–F49 on the homonyms); B_R = B9–B12 (33 claims). T against L_F(B_D) uses B1–B8.",
  "Census fields a fetched source contradicts or shows out of date (reported, not fixed): (1) operator, 'the saved source names no institution': the paper names Johns Hopkins University under the authors (B10, F23) and thanks the JHU HLTCOE; (2) holdings and lifecycle, 'a released live interface (unread)', 'whether the live interface is still served is unread': on 2026-11-01 dataportraits.org served its landing page (B2) and its three sketch front ends answered HTTP 502 (B12, F52); (3) count and holdings, 'two datasets documented (The Pile and The Stack)': the site lists a third sketch, The Stack V2 (F10); (4) stated_limits, 'the saved abstract states no limits': the paper states them (F30, F39, F40); (5) record.name, 'the saved abstract does not say whether a portrait keeps any author, title or source metadata': the paper says a Bloom filter stores only hashes of elements (F29) and that the approach cannot retrieve the context around a match (F39), its property 'Indexing' given up (F27, F28); (6) tll_reading quotes \"#186's 'is the work in the layer'\", a phrase in neither #186 nor #1543.",
  "Capture findings the field ledger shows wrong or needing qualification (reported, not fixed): (1) findings[1] gives 'lightweight' and the recording to 'The abstract'; the abstract is B_R, and on T's rail the artifact's card (B2, dataportraits.org) says 'minimal and efficient' and 'rapid, minimal' (F6, F7), never 'lightweight', which stands on no card; (2) findings[2], 'the abstract's authors … and its venue … are not named': the card's own page names both in its citation (F12), so they were on the rail's page as well; (3) findings[3], 'at an address the paste does not show': the card's page is https://dataportraits.org/ (its lead is the snippet verbatim), the live interface's landing page, served on 2026-11-01, its sketch front ends HTTP 502; (4) card 5's snippet carries 'between 1897 and 1990', recorded as rendered.",
  "Homonyms: three other entities of the name, carried as distinct entities (homonyms.distinct_entities); 'The Stack' is also the title of Benjamin Bratton's book, which the archive cites (#368, #833) and which the string pass caught; unadmitted.",
  "Not run: scripts/build_non.py (only --check), scripts/non_axn.py; panel.json not edited; the session seats the panel entry, mints the AXN (0BAD) and builds the pages."
 ],
 "composition": {
  "source": "this row (composed directly under the brief rebuild/negative-of-the-negative/v010/BRIEF-NEW-ROW-LAYER.md)",
  "composer": "Claude (Opus 5.5), subagent, 2026-11-01",
  "spec": "EA-NEGONT-02 v0.10 (#1726)",
  "defining_material_read": [
   {
    "what": "the specification v0.10",
    "file": "data/texts/AXN-0B79-text.md",
    "read": "§0–§11 whole (the procedure), Appendix headings"
   },
   {
    "what": "the briefs",
    "file": "rebuild/negative-of-the-negative/v010/BRIEF-NEW-ROW-LAYER.md; v010/BRIEF-NEW-ROW-AUTHORSHIP.md; v09/BRIEF-NEW-ROW-AGGREGATOR.md, BRIEF-NEW-ROW-LABOR.md, BRIEF-NEW-ROW-LEFT.md",
    "read": "whole"
   },
   {
    "what": "traininglayerliterature.org layer census record, row",
    "file": "/home/claude/fleet/traininglayerliterature-org/data/layer-census/census.json (rows[slug == 'data-portraits'], organizing_question, admission_rule, areas, anchor, schema_note); data/layer-census/rows/data-portraits.py",
    "read": "whole"
   },
   {
    "what": "census sources",
    "file": "/home/claude/fleet/traininglayerliterature-org/data/layer-census/sources/data-portraits--arxiv.*, manifest.json",
    "read": "whole"
   },
   {
    "what": "the capture",
    "file": "data/EA-WG-CAPTURES-01.json entries[slug == data-portraits-aio-20261101]; rebuild/capture-registry/intake-20261101-layer-aggregators/parts/data-portraits.json",
    "read": "whole (transcript read for T and the delta only)"
   },
   {
    "what": "field texts",
    "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1–B12",
    "read": "whole (B10 to §7 and the references' head; B1, B3, B4, B8 at the passages on their projects)"
   },
   {
    "what": "#186 Who Is Writing for the Machines?",
    "file": "data/texts/AXN-032C-text.md",
    "read": "whole"
   },
   {
    "what": "#1543 Generative Uptake",
    "file": "data/texts/AXN-0642-text.md",
    "read": "§1–§4 and §10 whole, the rest by its argument"
   },
   {
    "what": "#127, #1075, #837, #623, #939, #727, #264",
    "file": "data/texts/AXN-02B2, data/deposits/AXN-0444.md, data/texts/AXN-0352, -01DA, -03B7, -0278, -0042",
    "read": "at the candidate loci and their sections"
   },
   {
    "what": "the exemplar rows",
    "file": "datasets/negative-of-the-negative/v2/rows/viaf.json, datasets/negative-of-the-negative/v2/rows/eclipse.json (the collision exemplar), datasets/negative-of-the-negative/v2/rows/megadigital.json",
    "read": "structure field by field"
   }
  ],
  "cuts": [],
  "notes": "equal case; KO 40 sentences, 52 of 52 field and 65 of 65 archive claims cited; P body 303 words"
 },
 "composition_v09": {
  "source": "this row (composed directly under the brief rebuild/negative-of-the-negative/v010/BRIEF-NEW-ROW-LAYER.md)",
  "composer": "Claude (Opus 5.5), subagent, 2026-11-01",
  "spec": "EA-NEGONT-02 v0.10 (#1726)",
  "defining_material_read": [
   {
    "what": "the specification v0.10",
    "file": "data/texts/AXN-0B79-text.md",
    "read": "§0–§11 whole (the procedure), Appendix headings"
   },
   {
    "what": "the briefs",
    "file": "rebuild/negative-of-the-negative/v010/BRIEF-NEW-ROW-LAYER.md; v010/BRIEF-NEW-ROW-AUTHORSHIP.md; v09/BRIEF-NEW-ROW-AGGREGATOR.md, BRIEF-NEW-ROW-LABOR.md, BRIEF-NEW-ROW-LEFT.md",
    "read": "whole"
   },
   {
    "what": "traininglayerliterature.org layer census record, row",
    "file": "/home/claude/fleet/traininglayerliterature-org/data/layer-census/census.json (rows[slug == 'data-portraits'], organizing_question, admission_rule, areas, anchor, schema_note); data/layer-census/rows/data-portraits.py",
    "read": "whole"
   },
   {
    "what": "census sources",
    "file": "/home/claude/fleet/traininglayerliterature-org/data/layer-census/sources/data-portraits--arxiv.*, manifest.json",
    "read": "whole"
   },
   {
    "what": "the capture",
    "file": "data/EA-WG-CAPTURES-01.json entries[slug == data-portraits-aio-20261101]; rebuild/capture-registry/intake-20261101-layer-aggregators/parts/data-portraits.json",
    "read": "whole (transcript read for T and the delta only)"
   },
   {
    "what": "field texts",
    "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1–B12",
    "read": "whole (B10 to §7 and the references' head; B1, B3, B4, B8 at the passages on their projects)"
   },
   {
    "what": "#186 Who Is Writing for the Machines?",
    "file": "data/texts/AXN-032C-text.md",
    "read": "whole"
   },
   {
    "what": "#1543 Generative Uptake",
    "file": "data/texts/AXN-0642-text.md",
    "read": "§1–§4 and §10 whole, the rest by its argument"
   },
   {
    "what": "#127, #1075, #837, #623, #939, #727, #264",
    "file": "data/texts/AXN-02B2, data/deposits/AXN-0444.md, data/texts/AXN-0352, -01DA, -03B7, -0278, -0042",
    "read": "at the candidate loci and their sections"
   },
   {
    "what": "the exemplar rows",
    "file": "datasets/negative-of-the-negative/v2/rows/viaf.json, datasets/negative-of-the-negative/v2/rows/eclipse.json (the collision exemplar), datasets/negative-of-the-negative/v2/rows/megadigital.json",
    "read": "structure field by field"
   }
  ],
  "cuts": [],
  "notes": "equal case; KO 40 sentences, 52 of 52 field and 65 of 65 archive claims cited; P body 303 words"
 },
 "field_by_origin": {
  "cards": [
   "B2"
  ],
  "homonym_cards": [
   {
    "id": "B1",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 1 (another entity of the name)",
    "card": "giorgialupi — Data Portraits at TED — giorgialupi",
    "url": "https://giorgialupi.com/data-portraits-at-ted2017",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Giorgia Lupi / Accurat, Data Portraits at TED 2017",
    "snippet": "Project: Data Portraits at TED 2017 (Vancouver) by Accurat, commissioned by TED and David Stark Design for Target's social space. Concept: Personal questionnair...",
    "files": {
     "B1.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.html",
      "sha256": "2d8c62c8abf433dba538700415ac1e4f16bc4f2e5434869dc0415b776be6dd40"
     },
     "B1.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.txt",
      "sha256": "05ffdf5ab82d444b0c7a6176a0b7c91e2627b9a11ec3226e4573c7856e2666bd"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:25Z",
     "http": "200",
     "url": "https://giorgialupi.com/data-portraits-at-ted2017",
     "final_url": "https://giorgialupi.com/data-portraits-at-ted2017",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.html",
     "raw_sha256": "2d8c62c8abf433dba538700415ac1e4f16bc4f2e5434869dc0415b776be6dd40",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.txt"
     ],
     "note": "found by WebSearch on the card's title and site; the page carries the project; the snippet's 'Project: … Concept:' line is the card renderer's summary and is not on the page"
    },
    "surfaced_by_T": [
     "T1 card 1"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "Data Portraits at TED — giorgialupi",
    "card_snippet": "Project: Data Portraits at TED 2017 (Vancouver) by Accurat, commissioned by TED and David Stark Design for Target's social space. Concept: Personal questionnair..."
   },
   {
    "id": "B3",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 3 (another entity of the name)",
    "card": "Zinn Education Project — W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
    "url": "https://www.zinnedproject.org/materials/du-bois-data-portraits/",
    "kind": "Overview card of a homonym",
    "sense": "homonym: W. E. B. Du Bois's data portraits (1900) and the 2018 book",
    "snippet": "Book: W. E. B. Du Bois's Data Portraits: Visualizing Black America (2018, 144 pages, Princeton Architectural Press). Creators: W. E. B. Du Bois Center at Univer...",
    "files": {
     "B3.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.html",
      "sha256": "e26b7dbd030029048c6f8866ae42ac35386c190c06039ee42f7fd54a0737400a"
     },
     "B3.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.txt",
      "sha256": "e74023fb964fe6e5b721b8acac20fcdef9fa0bff47558e16a67f79cc8d5623ad"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:26Z",
     "http": "200",
     "url": "https://www.zinnedproject.org/materials/du-bois-data-portraits/",
     "final_url": "https://www.zinnedproject.org/materials/du-bois-data-portraits/",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.html",
     "raw_sha256": "e26b7dbd030029048c6f8866ae42ac35386c190c06039ee42f7fd54a0737400a",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.txt"
     ],
     "note": "found by WebSearch on the card's title and site; the page carries the snippet's book entry"
    },
    "surfaced_by_T": [
     "T1 card 3"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
    "card_snippet": "Book: W. E. B. Du Bois's Data Portraits: Visualizing Black America (2018, 144 pages, Princeton Architectural Press). Creators: W. E. B. Du Bois Center at Univer..."
   },
   {
    "id": "B4",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 4 (another entity of the name)",
    "card": "Pentagram — TED Data Portraits - Pentagram",
    "url": "https://www.pentagram.com/work/ted-data-portraits/story",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Giorgia Lupi, TED Data Portraits (Pentagram)",
    "snippet": "Recognizing what they had in common at a glance, attendees found it easy to start. Recognizing what they had in common at a glance, attendees found it easy to s...",
    "files": {
     "B4.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.html",
      "sha256": "e120a28cf59c714240bcd07aa7f894a7fd54bdbc8216acdfd98e7a0a37764b44"
     },
     "B4.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.txt",
      "sha256": "21031e22d13aff62ebb0e4ed7371e885f64db9d7d21112fece8b20cd1ec417ba"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:27Z",
     "http": "200",
     "url": "https://www.pentagram.com/work/ted-data-portraits/story",
     "final_url": "https://www.pentagram.com/work/ted-data-portraits/story",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.html",
     "raw_sha256": "e120a28cf59c714240bcd07aa7f894a7fd54bdbc8216acdfd98e7a0a37764b44",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.txt"
     ],
     "note": "found by WebSearch on the card's title and site; the page carries the snippet's caption"
    },
    "surfaced_by_T": [
     "T1 card 4"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "TED Data Portraits - Pentagram",
    "card_snippet": "Recognizing what they had in common at a glance, attendees found it easy to start. Recognizing what they had in common at a glance, attendees found it easy to s..."
   },
   {
    "id": "B5",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 5 (another entity of the name)",
    "card": "SparkMap — W.E.B. Du Bois & Data Visualization History - SparkMap",
    "url": null,
    "kind": "Overview card of a homonym",
    "sense": "homonym: W. E. B. Du Bois's data portraits (SparkMap)",
    "snippet": "Sociologist W.E.B. Du Bois created sixty hand-drawn \"data portraits\" between 1897 and 1990 with Atlanta University students and scholars, visualizing Black Amer...",
    "files": {},
    "fetched": {
     "census_copy": false,
     "result": "unreachable: no URL found; WebSearch on the card's title and site ('SparkMap \"W.E.B. Du Bois\" Data Visualization History'; 'sparkmap.org W.E.B. Du Bois sixty hand-drawn \"data portraits\" Atlanta University'; 'SparkMap blog Du Bois data visualization history') returned other pages on Du Bois's charts and no SparkMap page; not fetched; the card's title and snippet its only text (§3.1(a))"
    },
    "surfaced_by_T": [
     "T1 card 5"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question. The snippet's 'between 1897 and 1990' is recorded as rendered.",
    "card_title": "W.E.B. Du Bois & Data Visualization History - SparkMap",
    "card_snippet": "Sociologist W.E.B. Du Bois created sixty hand-drawn \"data portraits\" between 1897 and 1990 with Atlanta University students and scholars, visualizing Black Amer..."
   },
   {
    "id": "B6",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 6 (another entity of the name; a video card, 5m)",
    "card": "YouTube·University of Massachusetts Amherst Libraries — W. E. B. Du Bois's Data Portraits: Visualizing Black America",
    "url": null,
    "kind": "video card of a homonym",
    "sense": "homonym: W. E. B. Du Bois's Data Portraits (a University of Massachusetts Amherst Libraries video)",
    "snippet": null,
    "files": {},
    "fetched": {
     "census_copy": false,
     "result": "video card: title only (no snippet on the rail); admitted through its title (§3.1(a)); a WebSearch on the title and channel found no video URL; not fetched"
    },
    "surfaced_by_T": [
     "T1 card 6"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "W. E. B. Du Bois's Data Portraits: Visualizing Black America",
    "card_snippet": null
   },
   {
    "id": "B7",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 7 (another entity of the name)",
    "card": "MIT Media Lab — Data Portraits - MIT Media Lab",
    "url": "https://www.media.mit.edu/projects/data-portraits/overview/",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Judith Donath's data portraits (MIT Media Lab)",
    "snippet": "Data portraits depict their subjects' accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, soci...",
    "files": {},
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:27Z",
     "http": "000",
     "url": "https://www.media.mit.edu/projects/data-portraits/overview/",
     "result": "unreachable: TLS failure ('SSL certificate problem: self-signed certificate in certificate chain') for https://www.media.mit.edu/projects/data-portraits/overview/ (15:14:27Z) and https://www.media.mit.edu/publications/data-portraits/ (15:14:28Z), both found by WebSearch; one retry of the first with the proxy's CA bundle named explicitly (15:14:42Z), the same failure; verification never disabled; the card's title and snippet its only text (§3.1(a)); fetch.log"
    },
    "surfaced_by_T": [
     "T1 card 7"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "Data Portraits - MIT Media Lab",
    "card_snippet": "Data portraits depict their subjects' accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, soci..."
   },
   {
    "id": "B8",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 8 (another entity of the name)",
    "card": "ARTnews.com — 10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ...",
    "url": "https://www.artnews.com/art-news/news/data-portraits-reveal-strange-details-2474/",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Judith Donath's data portraits (ARTnews, 30 June 2014)",
    "snippet": "Your tweets, your Instagrams, your Facebook status updates, your search history—it's all a part of your data portrait. And right now, by reading this article, y...",
    "files": {
     "B8.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.html",
      "sha256": "bf5fb12c784943a1e43c4d6edf2784d38d20b3e54ad705c47550d893699c6cc8"
     },
     "B8.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.txt",
      "sha256": "550b7a468950e851e6ab1167538757c9c94ef2d73f3bb3cf28c1c7cbd0dcca19"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:15:07Z",
     "http": "200",
     "url": "https://www.artnews.com/art-news/news/data-portraits-reveal-strange-details-2474/",
     "final_url": "https://www.artnews.com/art-news/news/data-portraits-reveal-strange-details-2474/",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.html",
     "raw_sha256": "bf5fb12c784943a1e43c4d6edf2784d38d20b3e54ad705c47550d893699c6cc8",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.txt"
     ],
     "note": "URL from ARTnews's tag page for Ben Durham (found by WebSearch; fetched as tries/artnews-tag.html), which links the article by its title; the page carries the snippet's opening"
    },
    "surfaced_by_T": [
     "T1 card 8"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ...",
    "card_snippet": "Your tweets, your Instagrams, your Facebook status updates, your search history—it's all a part of your data portrait. And right now, by reading this article, y..."
   }
  ],
  "own_pages": [
   {
    "id": "B9",
    "origin": "own_page",
    "pool": "B_R",
    "role": "the entity's own page (the arXiv abstract; the census's one source, data-portraits--arxiv)",
    "card": "arXiv — [2303.03919] Data Portraits: Recording Foundation Model Training Data",
    "url": "https://arxiv.org/abs/2303.03919",
    "kind": "own page (the paper's abstract page)",
    "sense": "the entity",
    "files": {
     "B9.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.html",
      "sha256": "0a499acc8164a2d26699162d1c8a5890c9bd74686f68d213fac02b5e5fc79b90"
     },
     "B9.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt",
      "sha256": "cc5ddcc8fb7f00626127990cb5a61811cc22c033a0c8f373a1017d3258aa396e"
     }
    },
    "fetched": {
     "census_copy": true,
     "src": "data-portraits--arxiv",
     "read": "2026-11-01T05:07:12Z",
     "http": "200",
     "url": "https://arxiv.org/abs/2303.03919",
     "final_url": "https://arxiv.org/abs/2303.03919",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.html",
     "raw_sha256": "0a499acc8164a2d26699162d1c8a5890c9bd74686f68d213fac02b5e5fc79b90",
     "census_sha256": "0a499acc8164a2d26699162d1c8a5890c9bd74686f68d213fac02b5e5fc79b90",
     "agrees": true,
     "copied_from": "/home/claude/fleet/traininglayerliterature-org/data/layer-census/sources/data-portraits--arxiv.html.gz (gunzipped) and data-portraits--arxiv.txt",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt"
     ]
    },
    "surfaced_by_T": [],
    "names_operator": "yes ('Authors: Marc Marone , Benjamin Van Durme')",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R: admitted by the operator's ruling of 2026-10-29 ('as against their wikipedia articles if they exist, their own pages, etc.'); not surfaced by T. The census copy is intact (sha256 matches the manifest), so it was copied and not refetched."
   },
   {
    "id": "B10",
    "origin": "own_page",
    "pool": "B_R",
    "role": "the entity's own text (the paper, arXiv v2 PDF, linked from the abstract page as 'View PDF')",
    "card": "arXiv — Data Portraits: Recording Foundation Model Training Data (PDF, v2)",
    "url": "https://arxiv.org/pdf/2303.03919v2",
    "kind": "own page (the paper, full text)",
    "sense": "the entity",
    "files": {
     "B10.pdf": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.pdf",
      "sha256": "13d6b6d27a507a4a3118572218548da674c27d697cab721cde8f10c84625a43e"
     },
     "B10.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt",
      "sha256": "2bb2d7bf089592374d511248dee3c209d58e605c07899aa098d884e4c02daaaa"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:30Z",
     "http": "200",
     "url": "https://arxiv.org/pdf/2303.03919v2",
     "final_url": "https://arxiv.org/pdf/2303.03919v2",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.pdf",
     "raw_sha256": "13d6b6d27a507a4a3118572218548da674c27d697cab721cde8f10c84625a43e",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt"
     ],
     "note": "text by pdftotext (reading order, no layout); quotes checked against it whitespace-normalised"
    },
    "surfaced_by_T": [],
    "names_operator": "yes ('Marc Marone Benjamin Van Durme Johns Hopkins University')",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R: the entity's own text, linked from its own abstract page (B9); read whole to §7 and the references' head, the census's unchecked item 'the paper's full text beyond the abstract'. Not surfaced by T."
   },
   {
    "id": "B12",
    "origin": "own_page",
    "pool": "B_R",
    "role": "the entity's own pages: the three sketch front ends the landing page links (pile., stack., stack-v2.dataportraits.org)",
    "card": "dataportraits.org — the Pile, Stack and Stack V2 sketch front ends (HTTP 502)",
    "url": "https://pile.dataportraits.org/",
    "kind": "own pages (the live interface), unread: HTTP 502",
    "sense": "the entity",
    "files": {
     "B12-pile.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-pile.html",
      "sha256": "bb2fb7b51f18478e3ccde6376dc39285b20472f0394e7034fd2e0bdf6a5716e2"
     },
     "B12-stack.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-stack.html",
      "sha256": "bb2fb7b51f18478e3ccde6376dc39285b20472f0394e7034fd2e0bdf6a5716e2"
     },
     "B12-stack-v2.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-stack-v2.html",
      "sha256": "bb2fb7b51f18478e3ccde6376dc39285b20472f0394e7034fd2e0bdf6a5716e2"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:29Z / 15:15:08Z / 15:15:09Z",
     "http": "502",
     "urls": [
      "https://pile.dataportraits.org/",
      "https://stack.dataportraits.org/",
      "https://stack-v2.dataportraits.org/"
     ],
     "result": "unread: each answered HTTP 502 with the 16-byte body 'error code: 502'; one attempt each; recorded so and quoted only for that status"
    },
    "surfaced_by_T": [],
    "names_operator": "no",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R; the live interface the census records as 'unread'; not surfaced by T."
   }
  ],
  "wikipedia": [
   {
    "id": "B11",
    "origin": "wikipedia",
    "pool": "B_R",
    "role": "Wikipedia: the record of absence (no article under the name, its variants or the paper's title) and the search for the phrase",
    "card": "Wikipedia — no article 'Data Portraits' (HTTP 404); search for \"Data Portraits\"",
    "url": "https://en.wikipedia.org/w/index.php?title=Data_Portraits&action=raw",
    "kind": "encyclopedia (record of absence)",
    "sense": "the name",
    "files": {
     "B11-W-Data_Portraits.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits.html",
      "sha256": "6942501ce08d982a4a7d53dd2c13d4713528f385e9c95b9cbd44217c0058acd8"
     },
     "B11-W-Data_portraits.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portraits.html",
      "sha256": "e21e48c249870b2fd794da9a41ca0ab0c58eaaa2a10fbf8664ff569894b9dec6"
     },
     "B11-W-Data_portrait.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portrait.html",
      "sha256": "164d333b7e7c01c131d4d9e487fd5396a32c49a587e9a4bc057213d82011146a"
     },
     "B11-W-Data_Portraits_Recording_Foundation_Model_Training_Data.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits_Recording_Foundation_Model_Training_Data.html",
      "sha256": "9eab1c41eaaf26bfbfe08fc71e72ef323f467b85a6ba579b8d89181b9c763aeb"
     },
     "B11-search.json": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-search.json",
      "sha256": "39f19b42595c6af7c9838424d67f4e36f4d3a362caeb7d45f83524291a5ba9b3"
     },
     "B11.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11.txt",
      "sha256": "3daad4456f1fb998405a0801ee878dc1140cc9d7f9eb0f210ba8d0fda4e7c4ae"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "raw_wikitext": [
      {
       "title": "Data_Portraits",
       "http": "404",
       "read": "2026-11-01T15:14:31Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits.html",
       "sha256": "6942501ce08d982a4a7d53dd2c13d4713528f385e9c95b9cbd44217c0058acd8"
      },
      {
       "title": "Data_portraits",
       "http": "404",
       "read": "2026-11-01T15:14:32Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portraits.html",
       "sha256": "e21e48c249870b2fd794da9a41ca0ab0c58eaaa2a10fbf8664ff569894b9dec6"
      },
      {
       "title": "Data_portrait",
       "http": "404",
       "read": "2026-11-01T15:14:33Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portrait.html",
       "sha256": "164d333b7e7c01c131d4d9e487fd5396a32c49a587e9a4bc057213d82011146a"
      },
      {
       "title": "Data_Portraits:_Recording_Foundation_Model_Training_Data",
       "http": "404",
       "read": "2026-11-01T15:14:33Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits_Recording_Foundation_Model_Training_Data.html",
       "sha256": "9eab1c41eaaf26bfbfe08fc71e72ef323f467b85a6ba579b8d89181b9c763aeb"
      }
     ],
     "search": {
      "url": "https://en.wikipedia.org/w/api.php?action=query&list=search&srsearch=%22Data+Portraits%22&format=json&srlimit=50",
      "http": "200",
      "read": "2026-11-01T15:15:09Z",
      "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-search.json",
      "sha256": "39f19b42595c6af7c9838424d67f4e36f4d3a362caeb7d45f83524291a5ba9b3",
      "hits": "Dowell Myers; Infographic; The Exhibit of American Negroes; W. E. B. Du Bois; Jessica Helfand; Whitney Battle-Baptiste; Laurie Frick"
     },
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11.txt"
     ],
     "note": "B11.txt is the text of the first 404 page; its line of 'Sensitive client information' (the fetching host's IP address) is left out of the text, kept in the raw file"
    },
    "surfaced_by_T": [],
    "names_operator": "no",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R: the record of absence (v0.10 §3.1(a)); the census records no Wikipedia source for the entity."
   }
  ],
  "census_third_party": [],
  "unreachable": [
   {
    "id": "B5",
    "why": "no URL found by WebSearch; the snippet stands"
   },
   {
    "id": "B6",
    "why": "video card, title only"
   },
   {
    "id": "B7",
    "why": "TLS failure (self-signed certificate in chain) at both MIT Media Lab URLs; the snippet stands"
   },
   {
    "id": "B12",
    "why": "HTTP 502 at the three sketch front ends"
   },
   {
    "id": "B11",
    "why": "HTTP 404: no article (a record of absence)"
   }
  ],
  "title_only": [
   "B6"
  ],
  "homonym_note": "Seven of eight cards on T's rail (B1, B3–B8) are other entities of the name: Giorgia Lupi's Data Portraits at TED 2017 (B1, B4), W. E. B. Du Bois's data portraits and the 2018 book (B3, B5, B6), and Judith Donath's data portraits (B7, B8). In B_D as the cards are, each with its sense; their kernels ledgered (F43–F49) for the disambiguation and the first question.",
  "own_pages_note": "the census's one own page (data-portraits--arxiv, the abstract) is B9 (B_R); dataportraits.org, the entity's site, is card 2 (B2, B_D), carried once there; the paper (B10) and the sketch front ends (B12) added as the entity's own pages, the census's unchecked items",
  "wikipedia_state": "No article. Raw wikitext for 'Data_Portraits', 'Data_portraits', 'Data_portrait' and 'Data_Portraits:_Recording_Foundation_Model_Training_Data' each HTTP 404 (2026-11-01T15:14:31–33Z, agent 'Mozilla/5.0 (compatible; traininglayerliterature-census)'); the search API for \"Data Portraits\" returns seven articles (Dowell Myers; Infographic; The Exhibit of American Negroes; W. E. B. Du Bois; Jessica Helfand; Whitney Battle-Baptiste; Laurie Frick), the Du Bois book cited in several, none on the paper's artifact (B11, F50, F51). The census records no Wikipedia source for the entity.",
  "pools": {
   "B_D": [
    "B1",
    "B2",
    "B3",
    "B4",
    "B5",
    "B6",
    "B7",
    "B8"
   ],
   "B_R": [
    "B9",
    "B10",
    "B11",
    "B12"
   ],
   "note": "v0.10 §0.3, §3.1(a): B_D, every card on the rail (the body's inline links go to cards 2, 3 and 8, carried once as cards); B_R, the entity's own pages (the abstract, the paper, the sketch front ends) and the record of the encyclopedia's absence. T against L_F(B_D) uses B1–B8 alone."
  },
  "sources": [
   {
    "id": "B1",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 1 (another entity of the name)",
    "card": "giorgialupi — Data Portraits at TED — giorgialupi",
    "url": "https://giorgialupi.com/data-portraits-at-ted2017",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Giorgia Lupi / Accurat, Data Portraits at TED 2017",
    "snippet": "Project: Data Portraits at TED 2017 (Vancouver) by Accurat, commissioned by TED and David Stark Design for Target's social space. Concept: Personal questionnair...",
    "files": {
     "B1.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.html",
      "sha256": "2d8c62c8abf433dba538700415ac1e4f16bc4f2e5434869dc0415b776be6dd40"
     },
     "B1.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.txt",
      "sha256": "05ffdf5ab82d444b0c7a6176a0b7c91e2627b9a11ec3226e4573c7856e2666bd"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:25Z",
     "http": "200",
     "url": "https://giorgialupi.com/data-portraits-at-ted2017",
     "final_url": "https://giorgialupi.com/data-portraits-at-ted2017",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.html",
     "raw_sha256": "2d8c62c8abf433dba538700415ac1e4f16bc4f2e5434869dc0415b776be6dd40",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B1.txt"
     ],
     "note": "found by WebSearch on the card's title and site; the page carries the project; the snippet's 'Project: … Concept:' line is the card renderer's summary and is not on the page"
    },
    "surfaced_by_T": [
     "T1 card 1"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "Data Portraits at TED — giorgialupi",
    "card_snippet": "Project: Data Portraits at TED 2017 (Vancouver) by Accurat, commissioned by TED and David Stark Design for Target's social space. Concept: Personal questionnair..."
   },
   {
    "id": "B2",
    "origin": "card",
    "pool": "B_D",
    "role": "Overview card 2, the entity's own site (dataportraits.org); not a census source",
    "card": "Data Portraits — Data Portraits",
    "url": "https://dataportraits.org/",
    "kind": "the entity's home page (the live interface's landing page)",
    "sense": "the entity",
    "snippet": "Foundation models are trained on increasingly immense and opaque datasets. Even while these models are now key in AI system building, it can be difficult to ans...",
    "files": {
     "B2.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.html",
      "sha256": "5967547b30c0d39f01b55cb0306b1c355987047c8b46d0ceeec64aaf75efda15"
     },
     "B2.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt",
      "sha256": "492d654ff3e6621254fb319485bfc89f34e913a5606b25742786189af9123e6f"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:13:42Z",
     "http": "200",
     "url": "https://dataportraits.org/",
     "final_url": "https://dataportraits.org/",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.html",
     "raw_sha256": "5967547b30c0d39f01b55cb0306b1c355987047c8b46d0ceeec64aaf75efda15",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B2.txt"
     ],
     "note": "URL from the card's title and its snippet, which is the page's lead verbatim ('Foundation models are trained on increasingly immense and opaque datasets. Even while these models are now key in AI system building, it can be difficult to ans...'); the census did not save this page (its 'live interface … unread'); the arXiv abstract links it as 'this https URL' and the paper names it ('at dataportraits.org')"
    },
    "surfaced_by_T": [
     "T1 card 2",
     "T1 bullet 4 inline link"
    ],
    "names_operator": "yes (the BibTeX block: 'author = {Marone, Marc and {Van Durme}, Benjamin}')",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "The entity's own page, surfaced by T as card 2; carried once, in B_D (v0.10 §3.1(a)). Read whole.",
    "card_title": "Data Portraits",
    "card_snippet": "Foundation models are trained on increasingly immense and opaque datasets. Even while these models are now key in AI system building, it can be difficult to ans..."
   },
   {
    "id": "B3",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 3 (another entity of the name)",
    "card": "Zinn Education Project — W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
    "url": "https://www.zinnedproject.org/materials/du-bois-data-portraits/",
    "kind": "Overview card of a homonym",
    "sense": "homonym: W. E. B. Du Bois's data portraits (1900) and the 2018 book",
    "snippet": "Book: W. E. B. Du Bois's Data Portraits: Visualizing Black America (2018, 144 pages, Princeton Architectural Press). Creators: W. E. B. Du Bois Center at Univer...",
    "files": {
     "B3.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.html",
      "sha256": "e26b7dbd030029048c6f8866ae42ac35386c190c06039ee42f7fd54a0737400a"
     },
     "B3.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.txt",
      "sha256": "e74023fb964fe6e5b721b8acac20fcdef9fa0bff47558e16a67f79cc8d5623ad"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:26Z",
     "http": "200",
     "url": "https://www.zinnedproject.org/materials/du-bois-data-portraits/",
     "final_url": "https://www.zinnedproject.org/materials/du-bois-data-portraits/",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.html",
     "raw_sha256": "e26b7dbd030029048c6f8866ae42ac35386c190c06039ee42f7fd54a0737400a",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B3.txt"
     ],
     "note": "found by WebSearch on the card's title and site; the page carries the snippet's book entry"
    },
    "surfaced_by_T": [
     "T1 card 3"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "W. E. B. Du Bois's Data Portraits: Visualizing Black America - Zinn Education Project",
    "card_snippet": "Book: W. E. B. Du Bois's Data Portraits: Visualizing Black America (2018, 144 pages, Princeton Architectural Press). Creators: W. E. B. Du Bois Center at Univer..."
   },
   {
    "id": "B4",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 4 (another entity of the name)",
    "card": "Pentagram — TED Data Portraits - Pentagram",
    "url": "https://www.pentagram.com/work/ted-data-portraits/story",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Giorgia Lupi, TED Data Portraits (Pentagram)",
    "snippet": "Recognizing what they had in common at a glance, attendees found it easy to start. Recognizing what they had in common at a glance, attendees found it easy to s...",
    "files": {
     "B4.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.html",
      "sha256": "e120a28cf59c714240bcd07aa7f894a7fd54bdbc8216acdfd98e7a0a37764b44"
     },
     "B4.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.txt",
      "sha256": "21031e22d13aff62ebb0e4ed7371e885f64db9d7d21112fece8b20cd1ec417ba"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:27Z",
     "http": "200",
     "url": "https://www.pentagram.com/work/ted-data-portraits/story",
     "final_url": "https://www.pentagram.com/work/ted-data-portraits/story",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.html",
     "raw_sha256": "e120a28cf59c714240bcd07aa7f894a7fd54bdbc8216acdfd98e7a0a37764b44",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B4.txt"
     ],
     "note": "found by WebSearch on the card's title and site; the page carries the snippet's caption"
    },
    "surfaced_by_T": [
     "T1 card 4"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "TED Data Portraits - Pentagram",
    "card_snippet": "Recognizing what they had in common at a glance, attendees found it easy to start. Recognizing what they had in common at a glance, attendees found it easy to s..."
   },
   {
    "id": "B5",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 5 (another entity of the name)",
    "card": "SparkMap — W.E.B. Du Bois & Data Visualization History - SparkMap",
    "url": null,
    "kind": "Overview card of a homonym",
    "sense": "homonym: W. E. B. Du Bois's data portraits (SparkMap)",
    "snippet": "Sociologist W.E.B. Du Bois created sixty hand-drawn \"data portraits\" between 1897 and 1990 with Atlanta University students and scholars, visualizing Black Amer...",
    "files": {},
    "fetched": {
     "census_copy": false,
     "result": "unreachable: no URL found; WebSearch on the card's title and site ('SparkMap \"W.E.B. Du Bois\" Data Visualization History'; 'sparkmap.org W.E.B. Du Bois sixty hand-drawn \"data portraits\" Atlanta University'; 'SparkMap blog Du Bois data visualization history') returned other pages on Du Bois's charts and no SparkMap page; not fetched; the card's title and snippet its only text (§3.1(a))"
    },
    "surfaced_by_T": [
     "T1 card 5"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question. The snippet's 'between 1897 and 1990' is recorded as rendered.",
    "card_title": "W.E.B. Du Bois & Data Visualization History - SparkMap",
    "card_snippet": "Sociologist W.E.B. Du Bois created sixty hand-drawn \"data portraits\" between 1897 and 1990 with Atlanta University students and scholars, visualizing Black Amer..."
   },
   {
    "id": "B6",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 6 (another entity of the name; a video card, 5m)",
    "card": "YouTube·University of Massachusetts Amherst Libraries — W. E. B. Du Bois's Data Portraits: Visualizing Black America",
    "url": null,
    "kind": "video card of a homonym",
    "sense": "homonym: W. E. B. Du Bois's Data Portraits (a University of Massachusetts Amherst Libraries video)",
    "snippet": null,
    "files": {},
    "fetched": {
     "census_copy": false,
     "result": "video card: title only (no snippet on the rail); admitted through its title (§3.1(a)); a WebSearch on the title and channel found no video URL; not fetched"
    },
    "surfaced_by_T": [
     "T1 card 6"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "W. E. B. Du Bois's Data Portraits: Visualizing Black America",
    "card_snippet": null
   },
   {
    "id": "B7",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 7 (another entity of the name)",
    "card": "MIT Media Lab — Data Portraits - MIT Media Lab",
    "url": "https://www.media.mit.edu/projects/data-portraits/overview/",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Judith Donath's data portraits (MIT Media Lab)",
    "snippet": "Data portraits depict their subjects' accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, soci...",
    "files": {},
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:27Z",
     "http": "000",
     "url": "https://www.media.mit.edu/projects/data-portraits/overview/",
     "result": "unreachable: TLS failure ('SSL certificate problem: self-signed certificate in certificate chain') for https://www.media.mit.edu/projects/data-portraits/overview/ (15:14:27Z) and https://www.media.mit.edu/publications/data-portraits/ (15:14:28Z), both found by WebSearch; one retry of the first with the proxy's CA bundle named explicitly (15:14:42Z), the same failure; verification never disabled; the card's title and snippet its only text (§3.1(a)); fetch.log"
    },
    "surfaced_by_T": [
     "T1 card 7"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "Data Portraits - MIT Media Lab",
    "card_snippet": "Data portraits depict their subjects' accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, soci..."
   },
   {
    "id": "B8",
    "origin": "card (homonym)",
    "pool": "B_D",
    "role": "Overview card 8 (another entity of the name)",
    "card": "ARTnews.com — 10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ...",
    "url": "https://www.artnews.com/art-news/news/data-portraits-reveal-strange-details-2474/",
    "kind": "Overview card of a homonym",
    "sense": "homonym: Judith Donath's data portraits (ARTnews, 30 June 2014)",
    "snippet": "Your tweets, your Instagrams, your Facebook status updates, your search history—it's all a part of your data portrait. And right now, by reading this article, y...",
    "files": {
     "B8.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.html",
      "sha256": "bf5fb12c784943a1e43c4d6edf2784d38d20b3e54ad705c47550d893699c6cc8"
     },
     "B8.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.txt",
      "sha256": "550b7a468950e851e6ab1167538757c9c94ef2d73f3bb3cf28c1c7cbd0dcca19"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:15:07Z",
     "http": "200",
     "url": "https://www.artnews.com/art-news/news/data-portraits-reveal-strange-details-2474/",
     "final_url": "https://www.artnews.com/art-news/news/data-portraits-reveal-strange-details-2474/",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.html",
     "raw_sha256": "bf5fb12c784943a1e43c4d6edf2784d38d20b3e54ad705c47550d893699c6cc8",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B8.txt"
     ],
     "note": "URL from ARTnews's tag page for Ben Durham (found by WebSearch; fetched as tries/artnews-tag.html), which links the article by its title; the page carries the snippet's opening"
    },
    "surfaced_by_T": [
     "T1 card 8"
    ],
    "names_operator": "no (the entity's authors are not named)",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "In B_D as the card is (BRIEF-NEW-ROW-LAYER, substitution 7): recorded with its sense, its kernel ledgered for the disambiguation and the first question.",
    "card_title": "10 Painstakingly Crafted Data Portraits That Reveal Strange Details of People's ...",
    "card_snippet": "Your tweets, your Instagrams, your Facebook status updates, your search history—it's all a part of your data portrait. And right now, by reading this article, y..."
   },
   {
    "id": "B9",
    "origin": "own_page",
    "pool": "B_R",
    "role": "the entity's own page (the arXiv abstract; the census's one source, data-portraits--arxiv)",
    "card": "arXiv — [2303.03919] Data Portraits: Recording Foundation Model Training Data",
    "url": "https://arxiv.org/abs/2303.03919",
    "kind": "own page (the paper's abstract page)",
    "sense": "the entity",
    "files": {
     "B9.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.html",
      "sha256": "0a499acc8164a2d26699162d1c8a5890c9bd74686f68d213fac02b5e5fc79b90"
     },
     "B9.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt",
      "sha256": "cc5ddcc8fb7f00626127990cb5a61811cc22c033a0c8f373a1017d3258aa396e"
     }
    },
    "fetched": {
     "census_copy": true,
     "src": "data-portraits--arxiv",
     "read": "2026-11-01T05:07:12Z",
     "http": "200",
     "url": "https://arxiv.org/abs/2303.03919",
     "final_url": "https://arxiv.org/abs/2303.03919",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.html",
     "raw_sha256": "0a499acc8164a2d26699162d1c8a5890c9bd74686f68d213fac02b5e5fc79b90",
     "census_sha256": "0a499acc8164a2d26699162d1c8a5890c9bd74686f68d213fac02b5e5fc79b90",
     "agrees": true,
     "copied_from": "/home/claude/fleet/traininglayerliterature-org/data/layer-census/sources/data-portraits--arxiv.html.gz (gunzipped) and data-portraits--arxiv.txt",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B9.txt"
     ]
    },
    "surfaced_by_T": [],
    "names_operator": "yes ('Authors: Marc Marone , Benjamin Van Durme')",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R: admitted by the operator's ruling of 2026-10-29 ('as against their wikipedia articles if they exist, their own pages, etc.'); not surfaced by T. The census copy is intact (sha256 matches the manifest), so it was copied and not refetched."
   },
   {
    "id": "B10",
    "origin": "own_page",
    "pool": "B_R",
    "role": "the entity's own text (the paper, arXiv v2 PDF, linked from the abstract page as 'View PDF')",
    "card": "arXiv — Data Portraits: Recording Foundation Model Training Data (PDF, v2)",
    "url": "https://arxiv.org/pdf/2303.03919v2",
    "kind": "own page (the paper, full text)",
    "sense": "the entity",
    "files": {
     "B10.pdf": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.pdf",
      "sha256": "13d6b6d27a507a4a3118572218548da674c27d697cab721cde8f10c84625a43e"
     },
     "B10.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt",
      "sha256": "2bb2d7bf089592374d511248dee3c209d58e605c07899aa098d884e4c02daaaa"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:30Z",
     "http": "200",
     "url": "https://arxiv.org/pdf/2303.03919v2",
     "final_url": "https://arxiv.org/pdf/2303.03919v2",
     "raw": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.pdf",
     "raw_sha256": "13d6b6d27a507a4a3118572218548da674c27d697cab721cde8f10c84625a43e",
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B10.txt"
     ],
     "note": "text by pdftotext (reading order, no layout); quotes checked against it whitespace-normalised"
    },
    "surfaced_by_T": [],
    "names_operator": "yes ('Marc Marone Benjamin Van Durme Johns Hopkins University')",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R: the entity's own text, linked from its own abstract page (B9); read whole to §7 and the references' head, the census's unchecked item 'the paper's full text beyond the abstract'. Not surfaced by T."
   },
   {
    "id": "B11",
    "origin": "wikipedia",
    "pool": "B_R",
    "role": "Wikipedia: the record of absence (no article under the name, its variants or the paper's title) and the search for the phrase",
    "card": "Wikipedia — no article 'Data Portraits' (HTTP 404); search for \"Data Portraits\"",
    "url": "https://en.wikipedia.org/w/index.php?title=Data_Portraits&action=raw",
    "kind": "encyclopedia (record of absence)",
    "sense": "the name",
    "files": {
     "B11-W-Data_Portraits.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits.html",
      "sha256": "6942501ce08d982a4a7d53dd2c13d4713528f385e9c95b9cbd44217c0058acd8"
     },
     "B11-W-Data_portraits.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portraits.html",
      "sha256": "e21e48c249870b2fd794da9a41ca0ab0c58eaaa2a10fbf8664ff569894b9dec6"
     },
     "B11-W-Data_portrait.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portrait.html",
      "sha256": "164d333b7e7c01c131d4d9e487fd5396a32c49a587e9a4bc057213d82011146a"
     },
     "B11-W-Data_Portraits_Recording_Foundation_Model_Training_Data.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits_Recording_Foundation_Model_Training_Data.html",
      "sha256": "9eab1c41eaaf26bfbfe08fc71e72ef323f467b85a6ba579b8d89181b9c763aeb"
     },
     "B11-search.json": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-search.json",
      "sha256": "39f19b42595c6af7c9838424d67f4e36f4d3a362caeb7d45f83524291a5ba9b3"
     },
     "B11.txt": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11.txt",
      "sha256": "3daad4456f1fb998405a0801ee878dc1140cc9d7f9eb0f210ba8d0fda4e7c4ae"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "raw_wikitext": [
      {
       "title": "Data_Portraits",
       "http": "404",
       "read": "2026-11-01T15:14:31Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits.html",
       "sha256": "6942501ce08d982a4a7d53dd2c13d4713528f385e9c95b9cbd44217c0058acd8"
      },
      {
       "title": "Data_portraits",
       "http": "404",
       "read": "2026-11-01T15:14:32Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portraits.html",
       "sha256": "e21e48c249870b2fd794da9a41ca0ab0c58eaaa2a10fbf8664ff569894b9dec6"
      },
      {
       "title": "Data_portrait",
       "http": "404",
       "read": "2026-11-01T15:14:33Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_portrait.html",
       "sha256": "164d333b7e7c01c131d4d9e487fd5396a32c49a587e9a4bc057213d82011146a"
      },
      {
       "title": "Data_Portraits:_Recording_Foundation_Model_Training_Data",
       "http": "404",
       "read": "2026-11-01T15:14:33Z",
       "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-W-Data_Portraits_Recording_Foundation_Model_Training_Data.html",
       "sha256": "9eab1c41eaaf26bfbfe08fc71e72ef323f467b85a6ba579b8d89181b9c763aeb"
      }
     ],
     "search": {
      "url": "https://en.wikipedia.org/w/api.php?action=query&list=search&srsearch=%22Data+Portraits%22&format=json&srlimit=50",
      "http": "200",
      "read": "2026-11-01T15:15:09Z",
      "file": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11-search.json",
      "sha256": "39f19b42595c6af7c9838424d67f4e36f4d3a362caeb7d45f83524291a5ba9b3",
      "hits": "Dowell Myers; Infographic; The Exhibit of American Negroes; W. E. B. Du Bois; Jessica Helfand; Whitney Battle-Baptiste; Laurie Frick"
     },
     "text": [
      "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B11.txt"
     ],
     "note": "B11.txt is the text of the first 404 page; its line of 'Sensitive client information' (the fetching host's IP address) is left out of the text, kept in the raw file"
    },
    "surfaced_by_T": [],
    "names_operator": "no",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R: the record of absence (v0.10 §3.1(a)); the census records no Wikipedia source for the entity."
   },
   {
    "id": "B12",
    "origin": "own_page",
    "pool": "B_R",
    "role": "the entity's own pages: the three sketch front ends the landing page links (pile., stack., stack-v2.dataportraits.org)",
    "card": "dataportraits.org — the Pile, Stack and Stack V2 sketch front ends (HTTP 502)",
    "url": "https://pile.dataportraits.org/",
    "kind": "own pages (the live interface), unread: HTTP 502",
    "sense": "the entity",
    "files": {
     "B12-pile.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-pile.html",
      "sha256": "bb2fb7b51f18478e3ccde6376dc39285b20472f0394e7034fd2e0bdf6a5716e2"
     },
     "B12-stack.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-stack.html",
      "sha256": "bb2fb7b51f18478e3ccde6376dc39285b20472f0394e7034fd2e0bdf6a5716e2"
     },
     "B12-stack-v2.html": {
      "path": "datasets/negative-of-the-negative/v2/intake/data-portraits-20261101/B12-stack-v2.html",
      "sha256": "bb2fb7b51f18478e3ccde6376dc39285b20472f0394e7034fd2e0bdf6a5716e2"
     }
    },
    "fetched": {
     "census_copy": false,
     "agent": "Mozilla/5.0 (compatible; traininglayerliterature-census)",
     "read": "2026-11-01T15:14:29Z / 15:15:08Z / 15:15:09Z",
     "http": "502",
     "urls": [
      "https://pile.dataportraits.org/",
      "https://stack.dataportraits.org/",
      "https://stack-v2.dataportraits.org/"
     ],
     "result": "unread: each answered HTTP 502 with the 16-byte body 'error code: 502'; one attempt each; recorded so and quoted only for that status"
    },
    "surfaced_by_T": [],
    "names_operator": "no",
    "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, traininglayerliterature.org, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)",
    "note": "B_R; the live interface the census records as 'unread'; not surfaced by T."
   }
  ]
 },
 "derivatives": {
  "kernel": {
   "status": "composer-written",
   "derived_from": "the current composition, written by its composer under the brief",
   "note": "§8.2 derives the kernel mechanically from the frozen plan; this one was written by the composer, mechanical derivation pending."
  },
  "delta": {
   "status": "current",
   "derived_from": "the current composition"
  }
 }
}