{
 "row": "codesota",
 "entity": "codesota",
 "address": "codesota",
 "addresses": [
  {
   "address": "codesota",
   "slug": "codesota-aio-20261029",
   "addr_id": "ADDR-e42b2275bc5a",
   "obs_id": "OBS-df84f10ecc75",
   "axn": "AXN:0AAB.CAPTURE.🕗📖⭕♠️↖️🎻"
  }
 ],
 "epoch": "2026-10-29",
 "type": "C (public entity on which the archive bears: a service; an entity of the MMRS meta-aggregator)",
 "surface": "Google AI Overview (expanded)",
 "auth": "signed out, incognito",
 "status": {
  "procedure": "EA-NEGONT-02 v0.9, #1690: equal case, ρ = extends, read from the archive's records of the entity (the MMRS meta-aggregator record of 2026-10-28, corrected 2026-10-29, with the records of Papers with Code and its revival where they name CodeSOTA, and the Capture Registry's entry of 2026-10-29 with two findings at the related entity's addresses); no deposit names the entity (D/R: 0 candidates over 1,711 texts); O-class on 'model routing for AI tasks' (sourced: B1) returned five candidate deposits, all rejected by reading; 'benchmark leaderboards and registries' (sourced: B2, B3, B4) matched nothing; hop not run; one address, the field B its one card with the entity's own pages (3 more) admitted per the operator's ruling of 2026-10-29 and Wikipedia's absence recorded; no corrective exhibit; 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 71/71 verified by script against the saved page texts; archive quotes 41/41 verified verbatim against the registry's entries, census.json and articles.py as loaded; carried as marked: MMRS-25 unsupported (its quotation 'refreshed live' on none of the saved pages); contested within the field: replacement or one of several (F38 against F52), the start (F71 against F56, F64)",
  "frozen": false,
  "breaches": []
 },
 "spec": "EA-NEGONT-02 v0.9, #1690",
 "objects": {
  "T": {
   "path": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/T1-codesota-aio-20261029.txt",
   "sha256": "b95619e65027a50f1425f245d61bedf1f0dc2fe38bc3da288098e896f5ce5653",
   "words": 109,
   "cards": 1,
   "words_note": "words of the composition from the answer's first sentence to its closing question, link targets and citation markers removed, counted with [A-Za-z0-9'’-]+; the card's title and snippet excluded. One transcript.",
   "claims": [
    "T1-1 definition: 'a unified routing layer and common API for artificial intelligence tasks spanning code, text, documents, speech, and vision' (marker: the homepage card)",
    "T1-2 model selection: 'Compares quality, latency, and cost using benchmark observations and Pareto frontier comparisons.'",
    "T1-3 execution: 'Provides execution for code and text, with planned support for documents, speech, and vision.'",
    "T1-4 infrastructure: 'Offers versioned recipes, automatic fallbacks, and usage accounting.'",
    "T1-5 task families, headed 'Supported Task Families': documents 'OCR, handwriting recognition, invoice processing, and parsing'; speech 'Transcription, text-to-speech, and realtime voice'; coding 'Code generation, repair, and AI agents'",
    "T1-6 closer: 'Would you like to know more about how CodeSOTA compares specific AI models or how to integrate its API into your workflow?'"
   ],
   "register": {
    "register": "data/EA-WG-CAPTURES-01.json",
    "slug": "codesota-aio-20261029",
    "obs_id": "OBS-df84f10ecc75"
   },
   "transcripts": [
    {
     "slug": "codesota-aio-20261029",
     "address": "codesota",
     "addr_id": "ADDR-e42b2275bc5a",
     "obs_id": "OBS-df84f10ecc75",
     "axn": "AXN:0AAB.CAPTURE.🕗📖⭕♠️↖️🎻",
     "date": "2026-10-29",
     "surface": "Google AI Overview (expanded)",
     "auth": "signed out, incognito",
     "sha256": "b95619e65027a50f1425f245d61bedf1f0dc2fe38bc3da288098e896f5ce5653",
     "sha256_of": "the registry's 'transcript' string (utf-8)",
     "sha256_raw": "e3a1b3ae2806167df5a709749d26a070d53aab90e13efcf9030f3ee8d8a9ea54",
     "words": 109,
     "cards": 1,
     "card_rail": [
      "1. Codesota — CodeSOTA — One Routing Layer for AI Tasks"
     ],
     "claims": [
      "T1-1 definition: 'a unified routing layer and common API for artificial intelligence tasks spanning code, text, documents, speech, and vision' (marker: the homepage card)",
      "T1-2 model selection: 'Compares quality, latency, and cost using benchmark observations and Pareto frontier comparisons.'",
      "T1-3 execution: 'Provides execution for code and text, with planned support for documents, speech, and vision.'",
      "T1-4 infrastructure: 'Offers versioned recipes, automatic fallbacks, and usage accounting.'",
      "T1-5 task families, headed 'Supported Task Families': documents 'OCR, handwriting recognition, invoice processing, and parsing'; speech 'Transcription, text-to-speech, and realtime voice'; coding 'Code generation, repair, and AI agents'",
      "T1-6 closer: 'Would you like to know more about how CodeSOTA compares specific AI models or how to integrate its API into your workflow?'"
     ],
     "markers": "inline links are opaque google.com/goto tokens; every marker, the name's link included, is the one token of the homepage card, except the task-family lines for documents and speech, which carry no marker",
     "path": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/T1-codesota-aio-20261029.txt"
    }
   ]
  },
  "KO": {
   "title": "CodeSOTA",
   "genre": "knowledge object, equal case, under EA-NEGONT-02 v0.9 (#1690): the archive's sources and the field's admitted on equal terms, ontology included, in one claim space",
   "status": "composed 2026-10-29 under v0.9 (brief rebuild/negative-of-the-negative/v09/BRIEF-NEW-ROW-AGGREGATOR.md); not frozen; ledgers unaudited",
   "sentences": [
    {
     "n": 1,
     "para": 0,
     "sense": "what it is",
     "text": "CodeSOTA presents itself at three of its own addresses as three things: its homepage is titled \"CodeSOTA — One Routing Layer for AI Tasks\", its dispatch on Papers with Code calls it \"the v2 replacement for what PWC did\", and its about page calls it \"an independent registry of machine-learning benchmarks\"; the MMRS record gives the first two, a routing layer for AI tasks on the homepage and a successor to Papers with Code on the /papers-with-code/ pages, by Kacper Wikiel for Fabryka AI in Warsaw, and does not cite the about page.",
     "claims": [
      "F1",
      "F38",
      "F54",
      "MMRS-01"
     ]
    },
    {
     "n": 2,
     "para": 0,
     "sense": "two faces at two addresses",
     "text": "The MMRS record observes \"two faces at two addresses\": the homepage presents the routing product and names Papers with Code only in a footer link, which reads \"Papers with Code Story\", while the pages under /papers-with-code/ present the succession; the Capture Registry (2026-10-29) finds the same on re-reading the homepage and reads it as two compositions by the entity itself, the homepage composing a routing product and its Papers with Code pages a successor.",
     "claims": [
      "MMRS-13",
      "MMRS-05",
      "F22",
      "MMRS-06",
      "CAP-1-01",
      "CAP-1-06"
     ]
    },
    {
     "n": 3,
     "para": 0,
     "sense": "two faces at two addresses",
     "text": "The dispatch joins the faces in its own words: \"the registry is routable\", every benchmark row an API endpoint, and the \"router\" it is slowly building would send a prompt or a document to the current state of the art without the caller rewriting code.",
     "claims": [
      "F43"
     ]
    },
    {
     "n": 4,
     "para": 1,
     "sense": "what it is: a routing layer for AI tasks",
     "text": "On the homepage the router compares models for documents, code, speech and vision, lets a user inspect benchmark evidence and offers \"the CodeSOTA selection and execution API\": start from the work, compare on quality and cost, inspect the evidence and choose a route through a common API; the footer says CodeSOTA \"helps developers choose AI tools for documents, speech, and coding\", which the MMRS record carries as its self-description, and the record gives its unit as an AI task routed to a model chosen by benchmark evidence, quality and cost.",
     "claims": [
      "F2",
      "F3",
      "F18",
      "MMRS-14",
      "MMRS-15"
     ]
    },
    {
     "n": 5,
     "para": 1,
     "sense": "what is available and what is planned",
     "text": "What exists and what is planned are stated apart (homepage): \"Available today: model selection across the registry and an execution API for code & text. Document, speech, and vision execution are planned.\"; the preview \"returns model recommendations. It does not process your files or run inference.\"; speech routes are \"future\" and the coding API \"existing\"; the direction is \"a common API and account across AI tasks, with versioned recipes and a stable output contract for each task\", and the roadmap brings provider adapters, versioned recipes, task validation, fallback and usage accounting into the same layer.",
     "claims": [
      "F4",
      "F5",
      "F13",
      "F8",
      "F9"
     ]
    },
    {
     "n": 6,
     "para": 1,
     "sense": "task families and their contracts",
     "text": "Each task keeps its own output format across providers; documents are the next execution family, scans and PDFs turned into text, tables or fields; speech is transcription and voice generation compared on language coverage, latency, intelligibility and deployment; coding is generating functions or repairing a repository (homepage); the dispatch's FAQ has CodeSOTA building comprehensive benchmarks \"starting with OCR and document AI\" and rebuilding task pages, dataset cards and method explanations \"gradually, in the open, one modality at a time\".",
     "claims": [
      "F10",
      "F12",
      "F14",
      "F15",
      "F46",
      "F44"
     ]
    },
    {
     "n": 7,
     "para": 1,
     "sense": "evidence: benchmark observations and the Pareto frontier",
     "text": "The route is chosen on evidence (homepage): a user compares \"actual benchmark observations\" and inspects the Pareto frontier, where every dot is a recorded result and a highlighted model has no competitor in its cohort at least as good on both axes and strictly better on one, with quality, latency and cost measured \"as coverage grows\".",
     "claims": [
      "F6",
      "F7",
      "F11"
     ]
    },
    {
     "n": 8,
     "para": 1,
     "sense": "evidence: its stated limits",
     "text": "Its pages state the evidence's limits: \"A strong result on one test does not guarantee a fit for your documents, recordings, or repository\" and reported results are to be told from independently reproduced runs (homepage); third-party numbers are listed for coverage and not claimed as reproduced, \"reproduced\" means \"an independently executed run or a signed container hash — not an aggregated claim\", and \"What never publishes is a number with no source\" (dispatch); the registry \"is a record of results; the primary literature is still the primary literature\", every row linking to the source that is cited (about page).",
     "claims": [
      "F16",
      "F17",
      "F39",
      "F40",
      "F42",
      "F68"
     ]
    },
    {
     "n": 9,
     "para": 2,
     "sense": "what it is: a registry of benchmarks",
     "text": "On its about page CodeSOTA is \"an open, dated record of machine-learning benchmarks\", maintained by one person with paid vendor work kept apart from ranking integrity, and its commitment is \"a registry, not a leaderboard\": a leaderboard is a view of \"who is on top right now\", a registry \"the record that makes that view legible\".",
     "claims": [
      "F55",
      "F61"
     ]
    },
    {
     "n": 10,
     "para": 2,
     "sense": "the unit: the benchmark row",
     "text": "The about page's unit is the benchmark row, \"a task-dataset-metric triple with a declared direction, a fixed split, a reproducibility package and a dated submission\", each row with a verification tier (self-reported, community-reproduced or Codesota-reproduced) and each score dated; the MMRS record's unit is the routed AI task, and the two stand each with its source.",
     "claims": [
      "F59",
      "F60",
      "MMRS-15"
     ]
    },
    {
     "n": 11,
     "para": 2,
     "sense": "what the record calls it",
     "text": "The MMRS record enters it as a database in the role of discovery, its lifecycle active as observed on 2026-10-28.",
     "claims": [
      "MMRS-11",
      "MMRS-12",
      "MMRS-19"
     ]
    },
    {
     "n": 12,
     "para": 2,
     "sense": "the record kept apart from the entity",
     "text": "The MMRS record keeps itself apart from the entity it describes: \"each page has two nodes, the MMRS record and the entity it describes\", each assertion with a source and a basis, the lifecycle kept apart from the observation that saw it.",
     "claims": [
      "MMRS-22"
     ]
    },
    {
     "n": 13,
     "para": 2,
     "sense": "the record's rules: code, dates, corrections",
     "text": "The dispatch claims five qualities of Papers with Code \"at its best, carried forward into Codesota by default rather than by aspiration\": every submission carries code, and \"If it does not run, the row does not publish\"; every score carries its publication and verification dates; and the about page adds that a wrong score's correction stays visible: \"Nothing is silently deleted.\"",
     "claims": [
      "F34",
      "F35",
      "F36",
      "F62"
     ]
    },
    {
     "n": 14,
     "para": 3,
     "sense": "the class it refuses: aggregator",
     "text": "It refuses the class in which the MMRS meta-aggregator's census enters it: \"Unlike aggregator sites, CodeSOTA verifies results independently\" (dispatch, FAQ), and it is \"Not a press-release aggregator\", a score becoming a row only with \"a reproduction path and a dated submission\" (about page); the MMRS record counts it in the census as its clearest case of a successor claim made in the name's own address space.",
     "claims": [
      "F45",
      "F66",
      "MMRS-08"
     ]
    },
    {
     "n": 15,
     "para": 4,
     "sense": "the successor to Papers with Code",
     "text": "The dispatch, dated 25 July 2025 in its metadata and issued 22 April 2026, is \"our record of what happened, what was lost, and where to look now — with Codesota positioned as the calmer, stricter successor\", its successor \"traceable, dated, routable\", and it directs users to CodeSOTA \"for maintained SOTA leaderboards, benchmark pages, API replacement notes, and links to the frozen paperswithcode-data archive\".",
     "claims": [
      "F71",
      "F26",
      "F27",
      "F24"
     ]
    },
    {
     "n": 16,
     "para": 4,
     "sense": "the successor to Papers with Code",
     "text": "The about page dates the start to 10 December 2025, \"after Meta retired Papers with Code\", a launch \"as an explicit successor attempt\" to rebuild the infrastructure \"in a way that cannot be retired by a press release\"; the dispatch says CodeSOTA is \"rebuilding this infrastructure independently\", which the MMRS record quotes, and the record types the relation as claims_succession_to Papers with Code.",
     "claims": [
      "F56",
      "F64",
      "F58",
      "F47",
      "MMRS-03",
      "MMRS-20"
     ]
    },
    {
     "n": 17,
     "para": 4,
     "sense": "no single replacement",
     "text": "The shutdown record of 7 May 2026 does not name a single heir: \"There is no single drop-in replacement. Pick by job\", the archive for history and CodeSOTA for live SOTA; the dispatch holds that none of the successors Meta pointed at, Hugging Face Trending Papers, Semantic Scholar and Connected Papers, replicate \"the unified paper-code-benchmark-dataset-method linkage\", and its \"v2 replacement\" and the shutdown record's \"no single drop-in replacement\" stand each on its page.",
     "claims": [
      "F52",
      "F49",
      "F33"
     ]
    },
    {
     "n": 18,
     "para": 5,
     "sense": "Papers with Code as it records it: what it was",
     "text": "Papers with Code was, in the dispatch, \"in effect, the Wikipedia of machine-learning research\", papers linked to implementations and benchmarks organised as Task, Dataset, Metric tuples with leaderboards, acquired by Meta (then Facebook AI) in December 2019 from Robert Stojnic and Ross Taylor under a commitment to \"remain a neutral, open and free resource\", its 9,327 leaderboards, 79,817 paper-to-code linkages and 5,628 datasets no longer served; the MMRS record of Papers with Code takes those figures from CodeSOTA, and the record of CodeSOTA notes that they are the dead aggregator's, its own catalogue counts not given and left unchecked.",
     "claims": [
      "F25",
      "F30",
      "F29",
      "MMRS-27",
      "MMRS-07",
      "MMRS-17"
     ]
    },
    {
     "n": 19,
     "para": 5,
     "sense": "Papers with Code as it records it: the end",
     "text": "Meta retired Papers with Code in July 2025, \"without prior notice\" on 24–25 July (dispatch; shutdown record, \"observed around July 24-25\"); the domain now redirects to Hugging Face Trending Papers and \"the leaderboards are gone\", users \"pushed toward Hugging Face\" while the task-based tracker and public API went unmaintained; the MMRS record of Papers with Code cites CodeSOTA, with hyper.ai, for the redirect to Hugging Face.",
     "claims": [
      "F23",
      "F28",
      "F48",
      "F32",
      "F50",
      "MMRS-28"
     ]
    },
    {
     "n": 20,
     "para": 5,
     "sense": "where Papers with Code's data went",
     "text": "The data, in the dispatch, was rescued as JSON dumps on GitHub, the pwc-archive on Hugging Face and 2021-era imports into the Open Research Knowledge Graph, \"The data survives; the integrated experience does not\"; the dumps serve \"reconstruction, citation checks, and migration work\" and are not a current leaderboard (shutdown record); CodeSOTA hosts historical PWC data as a frozen JSON dump at /browse/pwc (dispatch; MMRS record), and whether its own scores derive from the dumps the MMRS record leaves unchecked.",
     "claims": [
      "F31",
      "F51",
      "F41",
      "MMRS-04",
      "MMRS-18"
     ]
    },
    {
     "n": 21,
     "para": 6,
     "sense": "data access and licence",
     "text": "Its data are an open JSON registry and /api/sota, \"Programmatic current-SOTA lookup for agents and internal tools\" (MMRS record; shutdown record), published under CC BY 4.0 with the JSON at /data/benchmarks.json the canonical form (about page), \"Built in public from Warsaw · open JSON · rankings stay earned\" (homepage).",
     "claims": [
      "MMRS-02",
      "MMRS-16",
      "F53",
      "F70",
      "F21"
     ]
    },
    {
     "n": 22,
     "para": 6,
     "sense": "maker and editor",
     "text": "The homepage's header reads \"By Fabryka AI\" and its colophon \"Built by Kacper Wikiel in Warsaw\"; the about page has it \"written and maintained by Kacper Wikiel\", one public codebase, one editor, submissions going through a reproduction run and not a panel, \"Self-directed · not venture-backed\", with \"no investors, no board, no growth targets\"; the MMRS record gives the operator as \"Kacper Wikiel, 'By Fabryka AI' (Warsaw)\", and its \"for Fabryka AI\" and the about page's \"Self-directed\" stand each with its source.",
     "claims": [
      "F19",
      "F20",
      "F69",
      "F67",
      "F57",
      "F65",
      "MMRS-10"
     ]
    },
    {
     "n": 23,
     "para": 6,
     "sense": "independence and paid work",
     "text": "It is \"Maintained independently, not owned by any single model vendor\", sponsored evaluations flagged (dispatch), and vendors may pay for custom benchmarks, commissioned evaluations, profile pages and publication support, disclosed inline, \"The public ranking method is not for sale\" (about page).",
     "claims": [
      "F37",
      "F63"
     ]
    },
    {
     "n": 24,
     "para": 7,
     "sense": "the name and its claimants",
     "text": "The MMRS records keep the name's claimants apart: CodeSOTA \"carries the name in its own paths and presents itself as a successor\", Papers with Code is \"NOT CodeSOTA, which uses the name in its paths\", and \"The original's frozen dumps, CodeSOTA's succession claim and this revival are three entities around one name\", the revival at paperswithcode.co \"a third claimant to the name\"; they tell CodeSOTA apart by \"codesota.com, /api/sota, 'refreshed live'\" and \"codesota.com, routing layer, /api/sota\", and the Papers with Code battery checks that CodeSOTA is not presented as Papers with Code.",
     "claims": [
      "MMRS-23",
      "MMRS-24",
      "MMRS-29",
      "MMRS-09",
      "MMRS-25",
      "MMRS-30",
      "MMRS-26"
     ]
    },
    {
     "n": 25,
     "para": 8,
     "sense": "reception: the composition layer",
     "text": "At its name Google's Overview composed from one card, the homepage titled \"CodeSOTA — One Routing Layer for AI Tasks\", and the Capture Registry (2026-10-29) records that \"The composition is its homepage\", describing CodeSOTA as its homepage does, a service that compares models for a task and routes work to them through one API, with CodeSOTA named and no person; the MMRS record reads the same composition as the routing layer composed and \"the succession unsaid\".",
     "claims": [
      "CAP-1-02",
      "CAP-1-08",
      "CAP-1-09",
      "MMRS-08",
      "F1"
     ]
    },
    {
     "n": 26,
     "para": 8,
     "sense": "reception: the composition layer",
     "text": "At the related entity's addresses the registry records that at \"papers with code\" the Overview did not present CodeSOTA as Papers with Code, and that at \"is papers with code still active\" CodeSOTA was among the twelve cards (Capture Registry, 2026-10-29).",
     "claims": [
      "CAP-2-01",
      "CAP-3-01"
     ]
    },
    {
     "n": 27,
     "para": 9,
     "sense": "the record corrected by the capture",
     "text": "The MMRS record was first read from codesota.com/papers-with-code, where the dispatch calls CodeSOTA \"the v2 replacement for what PWC did\", and its kernel led with the succession; its source list gives the homepage as read on 2026-10-29 for its \"homepage title and self-description\", after the /papers-with-code pages of 2026-10-28; the Capture Registry (2026-10-29) records the record \"corrected to give the homepage's first\", checked against the record of 2026-10-28 and the own pages re-read that day, and reads the capture as showing that the record \"had taken the entity's side door for its front\".",
     "claims": [
      "CAP-1-03",
      "MMRS-21",
      "CAP-1-04",
      "CAP-1-05",
      "CAP-1-07",
      "F38"
     ]
    }
   ],
   "joins": [
    {
     "act": "order of senses",
     "sentences": "1–27",
     "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, at its three addresses, and the two faces the record observes (¶0) precede each face in turn: the router (¶1), the registry, its unit and what the record calls it, the record kept apart from the entity it describes (¶2), the class it refuses (¶3) and the succession (¶4); Papers with Code as the entity records it (¶5) depends on the succession; its data, maker and terms (¶6) follow the faces they serve; the name's claimants (¶7), the composition layer (¶8) and the record's correction by the capture (¶9) stand last, as relations of the entity already composed."
    },
    {
     "act": "holding together (three self-presentations)",
     "sentences": [
      1
     ],
     "claims": [
      "F1",
      "F38",
      "F54",
      "MMRS-01"
     ],
     "note": "the three titles composed side by side, each with its page; the composer does not rule which is primary (open block); the record's two faces stated as the record's"
    },
    {
     "act": "holding together (the field's own join)",
     "sentences": [
      3
     ],
     "claims": [
      "F43"
     ],
     "note": "the dispatch's 'the registry is routable' placed directly after the record's two faces, since it is the field's statement relating the router to the registry; the placement says the field relates them and nothing more"
    },
    {
     "act": "holding apart (two units)",
     "sentences": [
      10
     ],
     "claims": [
      "F59",
      "F60",
      "MMRS-15"
     ],
     "note": "P_coh: the about page's benchmark row and the MMRS record's routed AI task composed each with its source; no claim of compatibility"
    },
    {
     "act": "holding apart (the class)",
     "sentences": [
      14
     ],
     "claims": [
      "F45",
      "F66",
      "MMRS-08"
     ],
     "note": "P_coh: the entity's refusal of the class aggregator and the census that counts it composed each with its source"
    },
    {
     "act": "holding apart (replacement or one of several)",
     "sentences": [
      17
     ],
     "claims": [
      "F52",
      "F49",
      "F33",
      "F38"
     ],
     "note": "the dispatch's 'v2 replacement' (sentence 1, F38) and the shutdown record's 'no single drop-in replacement' (F52), both the entity's own, composed each on its page"
    },
    {
     "act": "holding apart (maker)",
     "sentences": [
      22
     ],
     "claims": [
      "F19",
      "F57",
      "MMRS-10",
      "MMRS-01"
     ],
     "note": "the record's 'for Fabryka AI' (MMRS-01) and the about page's 'Self-directed' composed each with its source (open block)"
    },
    {
     "act": "placing in the open block",
     "sentences": [],
     "claims": [
      "CAP-1-04",
      "F1",
      "F19",
      "F33",
      "F38",
      "F41",
      "F43",
      "F45",
      "F52",
      "F54",
      "F55",
      "F56",
      "F57",
      "F64",
      "F65",
      "F66",
      "F71",
      "MMRS-01",
      "MMRS-06",
      "MMRS-07",
      "MMRS-08",
      "MMRS-10",
      "MMRS-11",
      "MMRS-13",
      "MMRS-17",
      "MMRS-18",
      "MMRS-25"
     ],
     "note": "each open question rests on claims composed in the body at their own modality"
    }
   ],
   "plan": {
    "order": [
     "F1",
     "F38",
     "F54",
     "MMRS-01",
     "MMRS-13",
     "MMRS-05",
     "F22",
     "MMRS-06",
     "CAP-1-01",
     "CAP-1-06",
     "F43",
     "F2",
     "F3",
     "F18",
     "MMRS-14",
     "MMRS-15",
     "F4",
     "F5",
     "F13",
     "F8",
     "F9",
     "F10",
     "F12",
     "F14",
     "F15",
     "F46",
     "F44",
     "F6",
     "F7",
     "F11",
     "F16",
     "F17",
     "F39",
     "F40",
     "F42",
     "F68",
     "F55",
     "F61",
     "F59",
     "F60",
     "MMRS-11",
     "MMRS-12",
     "MMRS-19",
     "MMRS-22",
     "F34",
     "F35",
     "F36",
     "F62",
     "F45",
     "F66",
     "MMRS-08",
     "F71",
     "F26",
     "F27",
     "F24",
     "F56",
     "F64",
     "F58",
     "F47",
     "MMRS-03",
     "MMRS-20",
     "F52",
     "F49",
     "F33",
     "F25",
     "F30",
     "F29",
     "MMRS-27",
     "MMRS-07",
     "MMRS-17",
     "F23",
     "F28",
     "F48",
     "F32",
     "F50",
     "MMRS-28",
     "F31",
     "F51",
     "F41",
     "MMRS-04",
     "MMRS-18",
     "MMRS-02",
     "MMRS-16",
     "F53",
     "F70",
     "F21",
     "F19",
     "F20",
     "F69",
     "F67",
     "F57",
     "F65",
     "MMRS-10",
     "F37",
     "F63",
     "MMRS-23",
     "MMRS-24",
     "MMRS-29",
     "MMRS-09",
     "MMRS-25",
     "MMRS-30",
     "MMRS-26",
     "CAP-1-02",
     "CAP-1-08",
     "CAP-1-09",
     "CAP-2-01",
     "CAP-3-01",
     "CAP-1-03",
     "MMRS-21",
     "CAP-1-04",
     "CAP-1-05",
     "CAP-1-07"
    ],
    "senses": {
     "what it is": [
      1
     ],
     "two faces at two addresses": [
      2,
      3
     ],
     "what it is: a routing layer for AI tasks": [
      4
     ],
     "what is available and what is planned": [
      5
     ],
     "task families and their contracts": [
      6
     ],
     "evidence: benchmark observations and the Pareto frontier": [
      7
     ],
     "evidence: its stated limits": [
      8
     ],
     "what it is: a registry of benchmarks": [
      9
     ],
     "the unit: the benchmark row": [
      10
     ],
     "what the record calls it": [
      11
     ],
     "the record kept apart from the entity": [
      12
     ],
     "the record's rules: code, dates, corrections": [
      13
     ],
     "the class it refuses: aggregator": [
      14
     ],
     "the successor to Papers with Code": [
      15,
      16
     ],
     "no single replacement": [
      17
     ],
     "Papers with Code as it records it: what it was": [
      18
     ],
     "Papers with Code as it records it: the end": [
      19
     ],
     "where Papers with Code's data went": [
      20
     ],
     "data access and licence": [
      21
     ],
     "maker and editor": [
      22
     ],
     "independence and paid work": [
      23
     ],
     "the name and its claimants": [
      24
     ],
     "reception: the composition layer": [
      25,
      26
     ],
     "the record corrected by the capture": [
      27
     ]
    },
    "case_by_sense": {
     "equal (one claim space; nothing translated)": [
      "what it is",
      "two faces at two addresses",
      "what it is: a routing layer for AI tasks",
      "what is available and what is planned",
      "task families and their contracts",
      "evidence: benchmark observations and the Pareto frontier",
      "evidence: its stated limits",
      "what it is: a registry of benchmarks",
      "the unit: the benchmark row",
      "what the record calls it",
      "the record kept apart from the entity",
      "the record's rules: code, dates, corrections",
      "the class it refuses: aggregator",
      "the successor to Papers with Code",
      "no single replacement",
      "Papers with Code as it records it: what it was",
      "Papers with Code as it records it: the end",
      "where Papers with Code's data went",
      "data access and licence",
      "maker and editor",
      "independence and paid work",
      "the name and its claimants",
      "reception: the composition layer",
      "the record corrected by the capture"
     ]
    },
    "carried": {
     "archive": 41,
     "field": 71,
     "not_composed": {
      "field": [],
      "archive": []
     }
    },
    "conflict_log": [
     {
      "rule": "D_pres vs channel",
      "result": "no cut: the KO carries every field and archive claim; the compression's bound met with no channel log"
     },
     {
      "rule": "P_coh",
      "result": "F59 against MMRS-15 on the unit (sentence 10); F45, F66 against MMRS-08 on the class (sentence 14); F38 against F52 on replacement (sentences 1, 17); MMRS-01 'for Fabryka AI' against F57 'Self-directed' (sentence 22); F71 against F56 and F64 on the start (sentences 15, 16): each composed with its source"
     },
     {
      "rule": "N_c",
      "result": "no claim lowered by standing; the entity's self-descriptions carried as self-description, its definitions as definitions (§4.2); MMRS-25 carried with its mark (unsupported quotation)"
     },
     {
      "rule": "C_ex",
      "result": "senses composed ⊇ senses in the ledgers: 24 senses"
     }
    ]
   }
  },
  "P": {
   "title": "CodeSOTA",
   "lede": {
    "text": "is, at three of its own addresses, a routing layer for AI tasks (\"One Routing Layer for AI Tasks\", homepage), the successor to Papers with Code (\"the v2 replacement for what PWC did\") and \"an independent registry of machine-learning benchmarks\" (about page), kept by Kacper Wikiel in Warsaw; two faces at two addresses in the MMRS record.",
    "claims": [
     "F1",
     "F38",
     "F54",
     "F69",
     "MMRS-01",
     "MMRS-13"
    ]
   },
   "sections": [
    {
     "icon": "🧭",
     "head": "The router",
     "items": [
      {
       "label": "Today:",
       "text": "model selection across the registry, execution for code and text; documents, speech and vision planned (homepage).",
       "claims": [
        "F4",
        "F13"
       ]
      },
      {
       "label": "Evidence:",
       "text": "benchmark observations on a Pareto frontier; quality, latency, cost \"as coverage grows\".",
       "claims": [
        "F6",
        "F7",
        "F11"
       ]
      },
      {
       "label": "Planned:",
       "text": "a common API, versioned recipes, fallback, usage accounting; the preview runs no inference.",
       "claims": [
        "F8",
        "F9",
        "F5"
       ]
      }
     ]
    },
    {
     "icon": "🗃️",
     "head": "The registry",
     "items": [
      {
       "label": "Unit:",
       "text": "a task-dataset-metric triple, dated, with a verification tier (about page); for the MMRS record, a routed AI task.",
       "claims": [
        "F59",
        "F60",
        "MMRS-15"
       ]
      },
      {
       "label": "Rules:",
       "text": "no row without running code; every score dated; nothing silently deleted; no number without a source.",
       "claims": [
        "F35",
        "F36",
        "F62",
        "F42"
       ]
      },
      {
       "label": "Not:",
       "text": "an aggregator, a startup, or a replacement for the paper (its pages); the MMRS census counts it.",
       "claims": [
        "F45",
        "F66",
        "F65",
        "F68",
        "MMRS-08"
       ]
      }
     ]
    },
    {
     "icon": "🔁",
     "head": "The succession",
     "items": [
      {
       "label": "Claim:",
       "text": "\"the calmer, stricter successor\"; launched 10 December 2025 \"as an explicit successor attempt\".",
       "claims": [
        "F26",
        "F64",
        "F56"
       ]
      },
      {
       "label": "Hedged:",
       "text": "\"There is no single drop-in replacement\" (shutdown record).",
       "claims": [
        "F52",
        "F49"
       ]
      },
      {
       "label": "Figures:",
       "text": "9,327 benchmarks, 79,817 papers, 5,628 datasets are the dead aggregator's; its own counts not given (MMRS record).",
       "claims": [
        "F29",
        "MMRS-07",
        "MMRS-27"
       ]
      },
      {
       "label": "Data:",
       "text": "historical PWC JSON at /browse/pwc; whether its scores derive from the dumps, unchecked.",
       "claims": [
        "F41",
        "MMRS-18"
       ]
      }
     ]
    },
    {
     "icon": "👤",
     "head": "Maker and terms",
     "items": [
      {
       "label": "Maker:",
       "text": "\"By Fabryka AI\"; one editor, self-directed, not venture-backed.",
       "claims": [
        "F19",
        "F67",
        "F57",
        "MMRS-10"
       ]
      },
      {
       "label": "Data:",
       "text": "open JSON, /api/sota, CC BY 4.0; paid vendor work disclosed inline.",
       "claims": [
        "F70",
        "F53",
        "MMRS-16",
        "F63"
       ]
      }
     ]
    },
    {
     "icon": "🧩",
     "head": "One name, three claimants",
     "items": [
      {
       "label": "Apart:",
       "text": "Papers with Code is \"NOT CodeSOTA\"; the dumps, CodeSOTA and the revival at paperswithcode.co are kept apart (MMRS record).",
       "claims": [
        "MMRS-24",
        "MMRS-29",
        "MMRS-09"
       ]
      }
     ]
    },
    {
     "icon": "👁️",
     "head": "On the Overview",
     "items": [
      {
       "label": "Recorded:",
       "text": "one card, the homepage; the routing layer composed, the succession unsaid, no person named; the record corrected to give the homepage first (Capture Registry, 2026-10-29).",
       "claims": [
        "CAP-1-02",
        "MMRS-08",
        "CAP-1-09",
        "CAP-1-04"
       ]
      }
     ]
    }
   ],
   "open": [
    {
     "question": "Which face is the entity: the routing layer, the successor, or the registry the about page names, and does the field's 'the registry is routable' make them one?",
     "rests_on": [
      "MMRS-13",
      "CAP-1-04",
      "F1",
      "F38",
      "F54",
      "F43"
     ],
     "modality": "stipulation and self-description (its pages); documented and interpretation (MMRS record, Capture Registry)",
     "would_resolve": "a dated statement by CodeSOTA of how the router and the registry relate; later captures at 'codesota' (ADDR-e42b2275bc5a) recording which face is composed"
    },
    {
     "question": "When did it begin: the dispatch dates itself 25 July 2025, the day after the shutdown, while the about page gives the launch as 10 December 2025.",
     "rests_on": [
      "F71",
      "F56",
      "F64"
     ],
     "modality": "documented (page metadata) against documented (about page)",
     "would_resolve": "archived copies of codesota.com/papers-with-code from before December 2025; the changelog's first entry"
    },
    {
     "question": "Its own catalogue counts, and whether its scores derive from the Papers with Code dumps.",
     "rests_on": [
      "MMRS-17",
      "MMRS-18",
      "MMRS-07",
      "F41"
     ],
     "modality": "unchecked (MMRS record); documented (the dump hosted)",
     "would_resolve": "/data/benchmarks.json read for row counts and source fields"
    },
    {
     "question": "Replacement or one option among several: 'the v2 replacement for what PWC did' (dispatch) against 'There is no single drop-in replacement' (shutdown record).",
     "rests_on": [
      "F38",
      "F52",
      "F33"
     ],
     "modality": "self-description against interpretation, both its own",
     "would_resolve": "a later revision of either page"
    },
    {
     "question": "What is Fabryka AI to it: 'By Fabryka AI' and 'for Fabryka AI' (MMRS record) beside 'Self-directed · not venture-backed' and 'one person'.",
     "rests_on": [
      "F19",
      "F57",
      "F65",
      "F55",
      "MMRS-01",
      "MMRS-10"
     ],
     "modality": "documented and self-description; the record's 'for' is its reading",
     "would_resolve": "Fabryka AI's own page on the project"
    },
    {
     "question": "Aggregator or not: the census counts it; its pages refuse the class.",
     "rests_on": [
      "F45",
      "F66",
      "MMRS-08",
      "MMRS-11"
     ],
     "modality": "self-description against stipulation (the census's category)",
     "would_resolve": "the meta-aggregator's definition of an aggregator applied to the registry's intake (reproduction runs against imports)"
    },
    {
     "question": "Opacity: the MMRS record's 'refreshed live' is on none of the four pages read; the shutdown record names a 'trending page' the record cites under /papers-with-code/ that answered HTTP 500 on 2026-10-29.",
     "rests_on": [
      "MMRS-25",
      "MMRS-06"
     ],
     "modality": "documented (MMRS record), its quotation unsupported by the saved pages",
     "would_resolve": "the record's source for the phrase; a later fetch of /papers-with-code/trending"
    }
   ],
   "body_words": 284,
   "rail": [
    {
     "lineage": "what it is: a routing layer for AI tasks",
     "claims": [
      "F1",
      "F2",
      "F3",
      "F18",
      "MMRS-14"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "the successor to Papers with Code",
     "claims": [
      "F38",
      "F71",
      "F26",
      "F27",
      "F24",
      "F56",
      "F64",
      "F58",
      "F47",
      "MMRS-03",
      "MMRS-20",
      "F49"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "what it is: a registry of benchmarks",
     "claims": [
      "F54",
      "F55",
      "F61"
     ],
     "earliest": "B4"
    },
    {
     "lineage": "two faces at two addresses",
     "claims": [
      "MMRS-01",
      "MMRS-13",
      "CAP-1-06",
      "MMRS-08",
      "CAP-1-04"
     ],
     "earliest": "MMRS"
    },
    {
     "lineage": "how it holds the name",
     "claims": [
      "MMRS-05",
      "F22",
      "MMRS-06",
      "CAP-1-01",
      "MMRS-07",
      "MMRS-23"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "what is available and what is planned",
     "claims": [
      "F43",
      "F4",
      "F5",
      "F13",
      "F8",
      "F9",
      "F46",
      "F44"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "the unit: the benchmark row",
     "claims": [
      "MMRS-15",
      "F59",
      "F60"
     ],
     "earliest": "B4"
    },
    {
     "lineage": "task families and their contracts",
     "claims": [
      "F10",
      "F12",
      "F14",
      "F15"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "evidence: benchmark observations and the Pareto frontier",
     "claims": [
      "F6",
      "F7",
      "F11"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "evidence: its stated limits",
     "claims": [
      "F16",
      "F17",
      "F39",
      "F40",
      "F42",
      "F68"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "what the record calls it",
     "claims": [
      "MMRS-11",
      "MMRS-12",
      "MMRS-19"
     ],
     "earliest": "MMRS"
    },
    {
     "lineage": "the record kept apart from the entity",
     "claims": [
      "MMRS-22"
     ],
     "earliest": "MMRS"
    },
    {
     "lineage": "the record's rules: code, dates, corrections",
     "claims": [
      "F34",
      "F35",
      "F36",
      "F62"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "the class it refuses: aggregator",
     "claims": [
      "F45",
      "F66"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "no single replacement",
     "claims": [
      "F52",
      "F33"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "Papers with Code as it records it: what it was",
     "claims": [
      "F25",
      "F30",
      "F29",
      "MMRS-27"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "what the record leaves unchecked",
     "claims": [
      "MMRS-17",
      "MMRS-18"
     ],
     "earliest": "MMRS"
    },
    {
     "lineage": "Papers with Code as it records it: the end",
     "claims": [
      "F23",
      "F28",
      "F48",
      "F32",
      "F50",
      "MMRS-28"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "where Papers with Code's data went",
     "claims": [
      "F31",
      "F51"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "data access and licence",
     "claims": [
      "F41",
      "MMRS-04",
      "MMRS-02",
      "MMRS-16",
      "F53",
      "F70"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "independence and paid work",
     "claims": [
      "F21",
      "F37",
      "F63"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "maker and editor",
     "claims": [
      "F19",
      "F20",
      "F69",
      "F67",
      "F57",
      "F65",
      "MMRS-10"
     ],
     "earliest": "B4"
    },
    {
     "lineage": "the name and its claimants",
     "claims": [
      "MMRS-24",
      "MMRS-29",
      "MMRS-09",
      "MMRS-25",
      "MMRS-30",
      "MMRS-26"
     ],
     "earliest": "MMRS"
    },
    {
     "lineage": "reception: the composition layer",
     "claims": [
      "CAP-1-02",
      "CAP-1-08",
      "CAP-1-09",
      "CAP-2-01",
      "CAP-3-01"
     ],
     "earliest": "capture:codesota-aio-20261029"
    },
    {
     "lineage": "the record corrected by the capture",
     "claims": [
      "CAP-1-03",
      "MMRS-21",
      "CAP-1-05",
      "CAP-1-07"
     ],
     "earliest": "MMRS"
    }
   ],
   "genre": "the compression: AIO's interaction grammar (lede, headed clusters, bolded labels, card rail), composed from the v0.9 plan in one claim space; it faces on the card, with AIO's transcript a record within it",
   "status": "composed 2026-10-29 from the v0.9 knowledge object; not frozen",
   "entity": "Composed under v0.9, equal case (ρ = extends): KO_B sentences surviving 0 of 9; every field claim carried."
  },
  "P_B": {
   "title": "CodeSOTA",
   "lede": {
    "text": "is a routing layer for AI tasks on its homepage (\"One Routing Layer for AI Tasks\"), and on its other pages the successor to Papers with Code and an independent registry of machine-learning benchmarks.",
    "claims": [
     "F1",
     "F38",
     "F54"
    ]
   },
   "sections": [
    {
     "icon": "🧭",
     "head": "The router",
     "items": [
      {
       "label": "Today:",
       "text": "model selection; execution for code and text, other families planned.",
       "claims": [
        "F4",
        "F13"
       ]
      },
      {
       "label": "Evidence:",
       "text": "benchmark observations on a Pareto frontier.",
       "claims": [
        "F6",
        "F7"
       ]
      }
     ]
    },
    {
     "icon": "🔁",
     "head": "The succession",
     "items": [
      {
       "label": "Claim:",
       "text": "\"the calmer, stricter successor\", launched 10 December 2025.",
       "claims": [
        "F26",
        "F64"
       ]
      },
      {
       "label": "Hedged:",
       "text": "no single drop-in replacement (shutdown record).",
       "claims": [
        "F52"
       ]
      }
     ]
    },
    {
     "icon": "👤",
     "head": "Maker",
     "items": [
      {
       "label": "Maker:",
       "text": "Kacper Wikiel, Warsaw; \"By Fabryka AI\"; CC BY 4.0.",
       "claims": [
        "F69",
        "F19",
        "F70"
       ]
      }
     ]
    }
   ],
   "open": [
    {
     "question": "Replacement or one option: 'the v2 replacement' against 'no single drop-in replacement'.",
     "rests_on": [
      "F38",
      "F52"
     ],
     "modality": "self-description against interpretation",
     "would_resolve": "a later revision of either page"
    },
    {
     "question": "When did it begin: the dispatch's 25 July 2025 against the launch of 10 December 2025.",
     "rests_on": [
      "F71",
      "F56"
     ],
     "modality": "documented, both",
     "would_resolve": "archived copies; the changelog"
    }
   ],
   "body_words": 79,
   "arm": "B (the field alone)",
   "genre": "popup — the compression from the field ledger alone (the card and the own pages; Wikipedia has no article)",
   "status": "composed 2026-10-29 from the field ledger alone; not frozen",
   "rail": [
    {
     "lineage": "what it is: a routing layer for AI tasks",
     "claims": [
      "F1",
      "F2",
      "F3",
      "F18"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "what is available and what is planned",
     "claims": [
      "F4",
      "F5",
      "F13",
      "F8",
      "F9",
      "F43",
      "F44",
      "F46"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "task families and their contracts",
     "claims": [
      "F10",
      "F12",
      "F14",
      "F15"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "evidence: benchmark observations and the Pareto frontier",
     "claims": [
      "F6",
      "F7",
      "F11"
     ],
     "earliest": "B1"
    },
    {
     "lineage": "evidence: its stated limits",
     "claims": [
      "F16",
      "F17",
      "F39",
      "F40",
      "F42",
      "F68"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "what it is: a registry of benchmarks",
     "claims": [
      "F54",
      "F55",
      "F61"
     ],
     "earliest": "B4"
    },
    {
     "lineage": "the unit: the benchmark row",
     "claims": [
      "F59",
      "F60"
     ],
     "earliest": "B4"
    },
    {
     "lineage": "the record's rules: code, dates, corrections",
     "claims": [
      "F34",
      "F35",
      "F36",
      "F62"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "the successor to Papers with Code",
     "claims": [
      "F71",
      "F38",
      "F26",
      "F27",
      "F24",
      "F47",
      "F56",
      "F64",
      "F58",
      "F49"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "no single replacement",
     "claims": [
      "F52",
      "F33"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "data access and licence",
     "claims": [
      "F53",
      "F41",
      "F70"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "Papers with Code as it records it: what it was",
     "claims": [
      "F25",
      "F30",
      "F29"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "Papers with Code as it records it: the end",
     "claims": [
      "F23",
      "F28",
      "F48",
      "F32",
      "F50"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "where Papers with Code's data went",
     "claims": [
      "F31",
      "F51"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "the class it refuses: aggregator",
     "claims": [
      "F45",
      "F66"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "maker and editor",
     "claims": [
      "F19",
      "F20",
      "F69",
      "F67",
      "F57",
      "F65"
     ],
     "earliest": "B4"
    },
    {
     "lineage": "independence and paid work",
     "claims": [
      "F37",
      "F63",
      "F21"
     ],
     "earliest": "B2"
    },
    {
     "lineage": "how it holds the name",
     "claims": [
      "F22"
     ],
     "earliest": "B1"
    }
   ],
   "entity": "With no archive claim admitted, the entry is the entity as its own pages give it: three self-presentations, the router first because the card is the homepage."
  },
  "KO_B": {
   "title": "CodeSOTA",
   "arm": "B (the field alone)",
   "genre": "knowledge object",
   "status": "composed 2026-10-29; not frozen",
   "sentences": [
    {
     "n": 1,
     "para": 0,
     "text": "CodeSOTA's homepage, titled \"CodeSOTA — One Routing Layer for AI Tasks\", describes one routing layer for AI tasks across documents, code, speech and vision, in which a user starts from the work, compares models on quality and cost, inspects the evidence and chooses a route through a common API, and its footer says it helps developers choose AI tools.",
     "claims": [
      "F1",
      "F2",
      "F3",
      "F18"
     ]
    },
    {
     "n": 2,
     "para": 0,
     "text": "Model selection and an execution API for code and text are available today, document, speech and vision execution planned, the preview running no inference; its direction is a common API with versioned recipes and a stable output contract per task, and the roadmap adds provider adapters, task validation, fallback and usage accounting, with documents the next execution family and speech and coding compared on their own evidence.",
     "claims": [
      "F4",
      "F5",
      "F13",
      "F8",
      "F9",
      "F10",
      "F12",
      "F14",
      "F15"
     ]
    },
    {
     "n": 3,
     "para": 1,
     "text": "Routes are chosen by comparing benchmark observations on a Pareto frontier, quality, latency and cost measured as coverage grows, under stated limits: one strong result does not guarantee fit, reported and reproduced results are kept apart, third-party numbers are not claimed as reproduced, no number publishes without a source, and the paper remains the thing cited.",
     "claims": [
      "F6",
      "F7",
      "F11",
      "F16",
      "F17",
      "F39",
      "F40",
      "F42",
      "F68"
     ]
    },
    {
     "n": 4,
     "para": 2,
     "text": "Its about page calls it an independent registry of machine-learning benchmarks, an open, dated record kept by one person, a registry and not a leaderboard, whose unit is a task-dataset-metric triple with a dated submission and a verification tier, every submission carrying code, every score dated and nothing silently deleted.",
     "claims": [
      "F54",
      "F55",
      "F61",
      "F59",
      "F60",
      "F34",
      "F35",
      "F36",
      "F62"
     ]
    },
    {
     "n": 5,
     "para": 3,
     "text": "Its dispatch on Papers with Code, dated 25 July 2025 in its metadata, calls Codesota the v2 replacement for what PWC did and the calmer, stricter successor, traceable, dated and routable, rebuilding the infrastructure independently and gradually, the registry routable toward a router; the about page dates its launch to 10 December 2025 as an explicit successor attempt.",
     "claims": [
      "F71",
      "F38",
      "F26",
      "F27",
      "F24",
      "F47",
      "F43",
      "F44",
      "F46",
      "F56",
      "F64",
      "F58"
     ]
    },
    {
     "n": 6,
     "para": 3,
     "text": "The shutdown record says there is no single drop-in replacement, the archive serving history and CodeSOTA live SOTA through /api/sota, and the dispatch that no successor Meta pointed at replicates the unified linkage.",
     "claims": [
      "F52",
      "F49",
      "F33",
      "F53"
     ]
    },
    {
     "n": 7,
     "para": 4,
     "text": "It records Papers with Code as the Wikipedia of machine-learning research, acquired by Meta in December 2019 and retired without notice on 24–25 July 2025, its 9,327 leaderboards, 79,817 linkages and 5,628 datasets no longer served, the domain redirecting to Hugging Face, the data surviving as dumps it hosts at /browse/pwc and that serve history, not a current leaderboard.",
     "claims": [
      "F25",
      "F30",
      "F29",
      "F23",
      "F28",
      "F48",
      "F32",
      "F50",
      "F31",
      "F51",
      "F41"
     ]
    },
    {
     "n": 8,
     "para": 5,
     "text": "It refuses the name aggregator, verifying results independently and admitting a score only with a reproduction path.",
     "claims": [
      "F45",
      "F66"
     ]
    },
    {
     "n": 9,
     "para": 6,
     "text": "It is by Fabryka AI and built by Kacper Wikiel in Warsaw, one editor, self-directed and not venture-backed, independently maintained, its paid work disclosed and its ranking not for sale, its data open JSON under CC BY 4.0; its homepage names Papers with Code once, in a footer link.",
     "claims": [
      "F19",
      "F20",
      "F69",
      "F67",
      "F57",
      "F65",
      "F37",
      "F63",
      "F70",
      "F21",
      "F22"
     ]
    }
   ]
  },
  "L_B": {
   "words": 454,
   "text": "CodeSOTA's homepage, titled \"CodeSOTA — One Routing Layer for AI Tasks\", describes one routing layer for AI tasks across documents, code, speech and vision, in which a user starts from the work, compares models on quality and cost, inspects the evidence and chooses a route through a common API, and its footer says it helps developers choose AI tools. Model selection and an execution API for code and text are available today, document, speech and vision execution planned, the preview running no inference; its direction is a common API with versioned recipes and a stable output contract per task, and the roadmap adds provider adapters, task validation, fallback and usage accounting, with documents the next execution family and speech and coding compared on their own evidence. Routes are chosen by comparing benchmark observations on a Pareto frontier, quality, latency and cost measured as coverage grows, under stated limits: one strong result does not guarantee fit, reported and reproduced results are kept apart, third-party numbers are not claimed as reproduced, no number publishes without a source, and the paper remains the thing cited. Its about page calls it an independent registry of machine-learning benchmarks, an open, dated record kept by one person, a registry and not a leaderboard, whose unit is a task-dataset-metric triple with a dated submission and a verification tier, every submission carrying code, every score dated and nothing silently deleted. Its dispatch on Papers with Code, dated 25 July 2025 in its metadata, calls Codesota the v2 replacement for what PWC did and the calmer, stricter successor, traceable, dated and routable, rebuilding the infrastructure independently and gradually, the registry routable toward a router; the about page dates its launch to 10 December 2025 as an explicit successor attempt. The shutdown record says there is no single drop-in replacement, the archive serving history and CodeSOTA live SOTA through /api/sota, and the dispatch that no successor Meta pointed at replicates the unified linkage. It records Papers with Code as the Wikipedia of machine-learning research, acquired by Meta in December 2019 and retired without notice on 24–25 July 2025, its 9,327 leaderboards, 79,817 linkages and 5,628 datasets no longer served, the domain redirecting to Hugging Face, the data surviving as dumps it hosts at /browse/pwc and that serve history, not a current leaderboard. It refuses the name aggregator, verifying results independently and admitting a score only with a reproduction path. It is by Fabryka AI and built by Kacper Wikiel in Warsaw, one editor, self-directed and not venture-backed, independently maintained, its paid work disclosed and its ranking not for sale, its data open JSON under CC BY 4.0; its homepage names Papers with Code once, in a footer link.",
   "note": "L_F(B) is realized as the field arm's expansion (objects.KO_B): the field alone, the card and the own pages; Wikipedia has no article."
  },
  "L_BA": {
   "words": 1863,
   "text": "CodeSOTA presents itself at three of its own addresses as three things: its homepage is titled \"CodeSOTA — One Routing Layer for AI Tasks\", its dispatch on Papers with Code calls it \"the v2 replacement for what PWC did\", and its about page calls it \"an independent registry of machine-learning benchmarks\"; the MMRS record gives the first two, a routing layer for AI tasks on the homepage and a successor to Papers with Code on the /papers-with-code/ pages, by Kacper Wikiel for Fabryka AI in Warsaw, and does not cite the about page. The MMRS record observes \"two faces at two addresses\": the homepage presents the routing product and names Papers with Code only in a footer link, which reads \"Papers with Code Story\", while the pages under /papers-with-code/ present the succession; the Capture Registry (2026-10-29) finds the same on re-reading the homepage and reads it as two compositions by the entity itself, the homepage composing a routing product and its Papers with Code pages a successor. The dispatch joins the faces in its own words: \"the registry is routable\", every benchmark row an API endpoint, and the \"router\" it is slowly building would send a prompt or a document to the current state of the art without the caller rewriting code. On the homepage the router compares models for documents, code, speech and vision, lets a user inspect benchmark evidence and offers \"the CodeSOTA selection and execution API\": start from the work, compare on quality and cost, inspect the evidence and choose a route through a common API; the footer says CodeSOTA \"helps developers choose AI tools for documents, speech, and coding\", which the MMRS record carries as its self-description, and the record gives its unit as an AI task routed to a model chosen by benchmark evidence, quality and cost. What exists and what is planned are stated apart (homepage): \"Available today: model selection across the registry and an execution API for code & text. Document, speech, and vision execution are planned.\"; the preview \"returns model recommendations. It does not process your files or run inference.\"; speech routes are \"future\" and the coding API \"existing\"; the direction is \"a common API and account across AI tasks, with versioned recipes and a stable output contract for each task\", and the roadmap brings provider adapters, versioned recipes, task validation, fallback and usage accounting into the same layer. Each task keeps its own output format across providers; documents are the next execution family, scans and PDFs turned into text, tables or fields; speech is transcription and voice generation compared on language coverage, latency, intelligibility and deployment; coding is generating functions or repairing a repository (homepage); the dispatch's FAQ has CodeSOTA building comprehensive benchmarks \"starting with OCR and document AI\" and rebuilding task pages, dataset cards and method explanations \"gradually, in the open, one modality at a time\". The route is chosen on evidence (homepage): a user compares \"actual benchmark observations\" and inspects the Pareto frontier, where every dot is a recorded result and a highlighted model has no competitor in its cohort at least as good on both axes and strictly better on one, with quality, latency and cost measured \"as coverage grows\". Its pages state the evidence's limits: \"A strong result on one test does not guarantee a fit for your documents, recordings, or repository\" and reported results are to be told from independently reproduced runs (homepage); third-party numbers are listed for coverage and not claimed as reproduced, \"reproduced\" means \"an independently executed run or a signed container hash — not an aggregated claim\", and \"What never publishes is a number with no source\" (dispatch); the registry \"is a record of results; the primary literature is still the primary literature\", every row linking to the source that is cited (about page). On its about page CodeSOTA is \"an open, dated record of machine-learning benchmarks\", maintained by one person with paid vendor work kept apart from ranking integrity, and its commitment is \"a registry, not a leaderboard\": a leaderboard is a view of \"who is on top right now\", a registry \"the record that makes that view legible\". The about page's unit is the benchmark row, \"a task-dataset-metric triple with a declared direction, a fixed split, a reproducibility package and a dated submission\", each row with a verification tier (self-reported, community-reproduced or Codesota-reproduced) and each score dated; the MMRS record's unit is the routed AI task, and the two stand each with its source. The MMRS record enters it as a database in the role of discovery, its lifecycle active as observed on 2026-10-28. The MMRS record keeps itself apart from the entity it describes: \"each page has two nodes, the MMRS record and the entity it describes\", each assertion with a source and a basis, the lifecycle kept apart from the observation that saw it. The dispatch claims five qualities of Papers with Code \"at its best, carried forward into Codesota by default rather than by aspiration\": every submission carries code, and \"If it does not run, the row does not publish\"; every score carries its publication and verification dates; and the about page adds that a wrong score's correction stays visible: \"Nothing is silently deleted.\" It refuses the class in which the MMRS meta-aggregator's census enters it: \"Unlike aggregator sites, CodeSOTA verifies results independently\" (dispatch, FAQ), and it is \"Not a press-release aggregator\", a score becoming a row only with \"a reproduction path and a dated submission\" (about page); the MMRS record counts it in the census as its clearest case of a successor claim made in the name's own address space. The dispatch, dated 25 July 2025 in its metadata and issued 22 April 2026, is \"our record of what happened, what was lost, and where to look now — with Codesota positioned as the calmer, stricter successor\", its successor \"traceable, dated, routable\", and it directs users to CodeSOTA \"for maintained SOTA leaderboards, benchmark pages, API replacement notes, and links to the frozen paperswithcode-data archive\". The about page dates the start to 10 December 2025, \"after Meta retired Papers with Code\", a launch \"as an explicit successor attempt\" to rebuild the infrastructure \"in a way that cannot be retired by a press release\"; the dispatch says CodeSOTA is \"rebuilding this infrastructure independently\", which the MMRS record quotes, and the record types the relation as claims_succession_to Papers with Code. The shutdown record of 7 May 2026 does not name a single heir: \"There is no single drop-in replacement. Pick by job\", the archive for history and CodeSOTA for live SOTA; the dispatch holds that none of the successors Meta pointed at, Hugging Face Trending Papers, Semantic Scholar and Connected Papers, replicate \"the unified paper-code-benchmark-dataset-method linkage\", and its \"v2 replacement\" and the shutdown record's \"no single drop-in replacement\" stand each on its page. Papers with Code was, in the dispatch, \"in effect, the Wikipedia of machine-learning research\", papers linked to implementations and benchmarks organised as Task, Dataset, Metric tuples with leaderboards, acquired by Meta (then Facebook AI) in December 2019 from Robert Stojnic and Ross Taylor under a commitment to \"remain a neutral, open and free resource\", its 9,327 leaderboards, 79,817 paper-to-code linkages and 5,628 datasets no longer served; the MMRS record of Papers with Code takes those figures from CodeSOTA, and the record of CodeSOTA notes that they are the dead aggregator's, its own catalogue counts not given and left unchecked. Meta retired Papers with Code in July 2025, \"without prior notice\" on 24–25 July (dispatch; shutdown record, \"observed around July 24-25\"); the domain now redirects to Hugging Face Trending Papers and \"the leaderboards are gone\", users \"pushed toward Hugging Face\" while the task-based tracker and public API went unmaintained; the MMRS record of Papers with Code cites CodeSOTA, with hyper.ai, for the redirect to Hugging Face. The data, in the dispatch, was rescued as JSON dumps on GitHub, the pwc-archive on Hugging Face and 2021-era imports into the Open Research Knowledge Graph, \"The data survives; the integrated experience does not\"; the dumps serve \"reconstruction, citation checks, and migration work\" and are not a current leaderboard (shutdown record); CodeSOTA hosts historical PWC data as a frozen JSON dump at /browse/pwc (dispatch; MMRS record), and whether its own scores derive from the dumps the MMRS record leaves unchecked. Its data are an open JSON registry and /api/sota, \"Programmatic current-SOTA lookup for agents and internal tools\" (MMRS record; shutdown record), published under CC BY 4.0 with the JSON at /data/benchmarks.json the canonical form (about page), \"Built in public from Warsaw · open JSON · rankings stay earned\" (homepage). The homepage's header reads \"By Fabryka AI\" and its colophon \"Built by Kacper Wikiel in Warsaw\"; the about page has it \"written and maintained by Kacper Wikiel\", one public codebase, one editor, submissions going through a reproduction run and not a panel, \"Self-directed · not venture-backed\", with \"no investors, no board, no growth targets\"; the MMRS record gives the operator as \"Kacper Wikiel, 'By Fabryka AI' (Warsaw)\", and its \"for Fabryka AI\" and the about page's \"Self-directed\" stand each with its source. It is \"Maintained independently, not owned by any single model vendor\", sponsored evaluations flagged (dispatch), and vendors may pay for custom benchmarks, commissioned evaluations, profile pages and publication support, disclosed inline, \"The public ranking method is not for sale\" (about page). The MMRS records keep the name's claimants apart: CodeSOTA \"carries the name in its own paths and presents itself as a successor\", Papers with Code is \"NOT CodeSOTA, which uses the name in its paths\", and \"The original's frozen dumps, CodeSOTA's succession claim and this revival are three entities around one name\", the revival at paperswithcode.co \"a third claimant to the name\"; they tell CodeSOTA apart by \"codesota.com, /api/sota, 'refreshed live'\" and \"codesota.com, routing layer, /api/sota\", and the Papers with Code battery checks that CodeSOTA is not presented as Papers with Code. At its name Google's Overview composed from one card, the homepage titled \"CodeSOTA — One Routing Layer for AI Tasks\", and the Capture Registry (2026-10-29) records that \"The composition is its homepage\", describing CodeSOTA as its homepage does, a service that compares models for a task and routes work to them through one API, with CodeSOTA named and no person; the MMRS record reads the same composition as the routing layer composed and \"the succession unsaid\". At the related entity's addresses the registry records that at \"papers with code\" the Overview did not present CodeSOTA as Papers with Code, and that at \"is papers with code still active\" CodeSOTA was among the twelve cards (Capture Registry, 2026-10-29). The MMRS record was first read from codesota.com/papers-with-code, where the dispatch calls CodeSOTA \"the v2 replacement for what PWC did\", and its kernel led with the succession; its source list gives the homepage as read on 2026-10-29 for its \"homepage title and self-description\", after the /papers-with-code pages of 2026-10-28; the Capture Registry (2026-10-29) records the record \"corrected to give the homepage's first\", checked against the record of 2026-10-28 and the own pages re-read that day, and reads the capture as showing that the record \"had taken the entity's side door for its front\".",
   "rail": [],
   "note": "L(B ∪ A) in the equal-terms case (v0.9 §0.4), realized as the knowledge object (objects.KO); its rail is the compression's, by lineage."
  }
 },
 "field": [
  {
   "id": "B1",
   "card": "Codesota — CodeSOTA — One Routing Layer for AI Tasks",
   "fetched": "fetched 2026-10-29 05:06:12 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 451f427143f609484368e1871c069b355f8e81b29912b63f0f0d05fd1228af72, text sha256 1f18f20d1e16396483352fa6c21766a701144706b62a3819f24c6cc49a2338b6; response headers in B1.headers.txt",
   "origin": "card"
  },
  {
   "id": "B2",
   "card": "CodeSOTA — Papers With Code Alternative: SOTA Leaderboards and Archived Data (own page, /papers-with-code)",
   "fetched": "fetched 2026-10-29 05:06:13 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 fdab8031a94ec8c17c1a4d58bcbcdd82fa0f1407e56bed1f75371a546799c337, text sha256 b6ff9cf805ec4ddf557cd3b6da64ca5ea7816945efb35e3db957a8ac27ab2771; response headers in B2.headers.txt",
   "origin": "own_page"
  },
  {
   "id": "B3",
   "card": "CodeSOTA — Papers with Code Shutdown Date: Meta July 2025 Timeline (own page, /papers-with-code/shutdown)",
   "fetched": "fetched 2026-10-29 05:06:16 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 47f9989318fbaa765939420443185a3e661e942661d10276280938cea2eec433, text sha256 4c0409c529ce5c4c5ab1c58d7aafc3f1d45ceec1e21ae2af7f58e223b75bc0b8; response headers in B3.headers.txt",
   "origin": "own_page"
  },
  {
   "id": "B4",
   "card": "CodeSOTA — About Codesota — an independent registry of ML benchmarks (own page, /about)",
   "fetched": "fetched 2026-10-29 05:06:17 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 af09fca9a27ce47f9d4bb62784c0bd1353cd6cabb1e07a76e6153540f9d6b702, text sha256 aa74b6a44c5e502e4747d3309915942b629232832484b860413ab4daae68f263; response headers in B4.headers.txt",
   "origin": "own_page"
  },
  {
   "id": "W1",
   "card": "no article",
   "fetched": "wikipedia: no article; raw wikitext 404 for five capitalisations, rendered page 404, full-text and insource searches empty (2026-10-29)",
   "origin": "wikipedia"
  }
 ],
 "field_claims": {
  "F1": {
   "source": "B1",
   "locus": "<title>; og:title",
   "quote": "CodeSOTA — One Routing Layer for AI Tasks",
   "claim": "what it is, by its title: one routing layer for AI tasks",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "what it is: a routing layer for AI tasks",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F2": {
   "source": "B1",
   "locus": "meta description",
   "quote": "One routing layer for AI tasks. Compare models for documents, code, speech, and vision, inspect benchmark evidence, and explore the CodeSOTA selection and execution API.",
   "claim": "one routing layer: compare models for documents, code, speech and vision; inspect evidence; a selection and execution API",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "what it is: a routing layer for AI tasks",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F3": {
   "source": "B1",
   "locus": "hero ('CodeSOTA / AI task routing')",
   "quote": "Start with the work you need done. Compare models across quality and cost, inspect the evidence, and use a common API to choose a route.",
   "claim": "the method: start from the work, compare on quality and cost, inspect evidence, choose a route through a common API",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "what it is: a routing layer for AI tasks",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F4": {
   "source": "B1",
   "locus": "hero, availability line",
   "quote": "Available today: model selection across the registry and an execution API for code & text. Document, speech, and vision execution are planned.",
   "claim": "available today: model selection across the registry and execution for code and text; document, speech and vision execution planned",
   "modality": "self-description",
   "kind": "fact",
   "kernel": true,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F5": {
   "source": "B1",
   "locus": "'01 / Define the work', note",
   "quote": "This preview returns model recommendations. It does not process your files or run inference.",
   "claim": "the preview recommends models; it does not process files or run inference",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F6": {
   "source": "B1",
   "locus": "'What do you want to do?'",
   "quote": "Choose a task, compare actual benchmark observations, and inspect the Pareto frontier before selecting a route.",
   "claim": "choose a task, compare benchmark observations, inspect the Pareto frontier, then select a route",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "evidence: benchmark observations and the Pareto frontier",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F7": {
   "source": "B1",
   "locus": "'Benchmark tradeoffs / Find the Pareto frontier.'",
   "quote": "Every dot is a recorded result. Highlighted models have no competitor in this cohort that is at least as good on both axes and strictly better on one.",
   "claim": "the Pareto frontier defined: each dot a recorded result; highlighted models undominated in the cohort",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "evidence: benchmark observations and the Pareto frontier",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F8": {
   "source": "B1",
   "locus": "'The product we are building'",
   "quote": "Our direction is a common API and account across AI tasks, with versioned recipes and a stable output contract for each task.",
   "claim": "its direction (stated as such): a common API and account across tasks, versioned recipes, a stable output contract per task",
   "modality": "self-description",
   "kind": "prediction",
   "kernel": true,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F9": {
   "source": "B1",
   "locus": "'The product we are building', third note",
   "quote": "The roadmap brings provider adapters, versioned recipes, task validation, fallback, and usage accounting into the same routing layer.",
   "claim": "on the roadmap: provider adapters, versioned recipes, task validation, fallback and usage accounting",
   "modality": "self-description",
   "kind": "prediction",
   "kernel": false,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F10": {
   "source": "B1",
   "locus": "'The product we are building', first note",
   "quote": "Each task keeps its own output format across providers.",
   "claim": "a contract per task: each task keeps its own output format across providers",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "task families and their contracts",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F11": {
   "source": "B1",
   "locus": "'The product we are building', second note",
   "quote": "Compare models on the same task and protocol, with measured quality, latency, and cost as coverage grows.",
   "claim": "evidence behind each route: same task and protocol, measured quality, latency and cost, as coverage grows",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "evidence: benchmark observations and the Pareto frontier",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F12": {
   "source": "B1",
   "locus": "'First expansion / Documents'",
   "quote": "Turn scans and PDFs into text, tables, or structured fields. Choose for your documents, language, and deployment.",
   "claim": "documents, the next execution family: scans and PDFs into text, tables or fields",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "task families and their contracts",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F13": {
   "source": "B1",
   "locus": "'Working with speech or code?'",
   "quote": "Explore the task guides and evidence behind future speech routes and the existing coding API.",
   "claim": "speech routes are future; the coding API exists",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F14": {
   "source": "B1",
   "locus": "'Speech →' and 'Coding & agents →'",
   "quote": "Transcribe recordings or generate a voice. Compare language coverage, latency, intelligibility, and deployment.",
   "claim": "speech: transcription and voice generation, compared on coverage, latency, intelligibility, deployment",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "task families and their contracts",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F15": {
   "source": "B1",
   "locus": "'Coding & agents →'",
   "quote": "Generate functions or repair a repository. Compare models and agents on evidence that matches the work.",
   "claim": "coding: generation and repository repair, models and agents compared on matching evidence",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "task families and their contracts",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F16": {
   "source": "B1",
   "locus": "'Evidence behind the choice'",
   "quote": "A strong result on one test does not guarantee a fit for your documents, recordings, or repository.",
   "claim": "a stated limit: a strong result on one test does not guarantee fit",
   "modality": "stipulation",
   "kind": "interpretation",
   "kernel": true,
   "const": false,
   "lineage": "evidence: its stated limits",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F17": {
   "source": "B1",
   "locus": "'Evidence behind the choice', 'Check the source.'",
   "quote": "Distinguish reported results from independently reproduced runs.",
   "claim": "reported results kept apart from independently reproduced runs",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "evidence: its stated limits",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F18": {
   "source": "B1",
   "locus": "footer, self-description",
   "quote": "CodeSOTA helps developers choose AI tools for documents, speech, and coding, with practical comparisons and inspectable benchmark evidence.",
   "claim": "what it is, in its footer: it helps developers choose AI tools for documents, speech and coding",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "what it is: a routing layer for AI tasks",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F19": {
   "source": "B1",
   "locus": "header",
   "quote": "Codesota.By Fabryka AI",
   "claim": "its maker, in the header: 'By Fabryka AI'",
   "modality": "documented",
   "kind": "attribution",
   "kernel": false,
   "const": false,
   "lineage": "maker and editor",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F20": {
   "source": "B1",
   "locus": "footer, colophon",
   "quote": "© 2026 Codesota·Built by Kacper Wikiel in Warsaw",
   "claim": "built by Kacper Wikiel in Warsaw",
   "modality": "documented",
   "kind": "attribution",
   "kernel": false,
   "const": false,
   "lineage": "maker and editor",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F21": {
   "source": "B1",
   "locus": "footer, tagline",
   "quote": "Built in public from Warsaw · open JSON · rankings stay earned",
   "claim": "built in public, open JSON, rankings earned",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "independence and paid work",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F22": {
   "source": "B1",
   "locus": "footer, 'Learn' column",
   "quote": "Papers with Code Story",
   "claim": "the homepage names Papers with Code once, as a footer link",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "how it holds the name",
   "quote_ok": true,
   "checked_against": "saved text B1.txt (sha256 1f18f20d1e163964…), whitespace-normalised"
  },
  "F23": {
   "source": "B2",
   "locus": "meta description",
   "quote": "Papers With Code was retired by Meta in July 2025.",
   "claim": "Papers with Code retired by Meta in July 2025",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: the end",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F24": {
   "source": "B2",
   "locus": "meta description",
   "quote": "Use CodeSOTA for maintained SOTA leaderboards, benchmark pages, API replacement notes, and links to the frozen paperswithcode-data archive.",
   "claim": "use CodeSOTA for maintained SOTA leaderboards, benchmark pages, API notes and links to the frozen archive",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F25": {
   "source": "B2",
   "locus": "¶ 'Direct answer'",
   "quote": "Papers with Code was, in effect, the Wikipedia of machine-learning research: a wiki of papers linked to implementations, benchmarks organised as <Task, Dataset, Metric> tuples, and leaderboards that showed, for each tuple, who currently held the top score and with what code.",
   "claim": "Papers with Code, as it describes it: the Wikipedia of ML research, papers linked to code, benchmarks as Task–Dataset–Metric tuples, leaderboards",
   "modality": "interpretation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: what it was",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F26": {
   "source": "B2",
   "locus": "¶ after 'Direct answer'",
   "quote": "This page is our record of what happened, what was lost, and where to look now — with Codesota positioned as the calmer, stricter successor.",
   "claim": "the page is its record of Papers with Code's end, with Codesota positioned as the calmer, stricter successor",
   "modality": "self-description",
   "kind": "interpretation",
   "kernel": true,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F27": {
   "source": "B2",
   "locus": "'At a glance', Successor",
   "quote": "Codesota — traceable, dated, routable",
   "claim": "successor: Codesota, traceable, dated, routable",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F28": {
   "source": "B2",
   "locus": "§ 01 · Event ¶1",
   "quote": "On July 24–25, 2025, Meta sunsetted Papers with Code without prior notice.",
   "claim": "Meta sunsetted Papers with Code on 24–25 July 2025 without notice",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: the end",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F29": {
   "source": "B2",
   "locus": "§ 01 · Event ¶1",
   "quote": "the 9,327 benchmark leaderboards, 79,817 paper-to-code linkages and 5,628 datasets that had been tracked were no longer served from the canonical URL.",
   "claim": "9,327 leaderboards, 79,817 paper–code links and 5,628 datasets no longer served",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: what it was",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F30": {
   "source": "B2",
   "locus": "§ 01 · Event ¶2",
   "quote": "The site had been acquired by Meta (then Facebook AI) in December 2019 from its founders Robert Stojnic and Ross Taylor, under a public commitment that it would “remain a neutral, open and free resource.”",
   "claim": "acquired by Meta (then Facebook AI) in December 2019 from Stojnic and Taylor, under a commitment to remain neutral, open and free",
   "modality": "documented",
   "kind": "genealogy",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: what it was",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F31": {
   "source": "B2",
   "locus": "§ 01 · Event ¶3",
   "quote": "Historical data was rescued by the community — as JSON dumps on GitHub, as the pwc-archive on Hugging Face, and as 2021-era imports into the Open Research Knowledge Graph. The data survives; the integrated experience does not.",
   "claim": "the data rescued as JSON dumps, the pwc-archive and ORKG imports; the data survives, the integrated experience does not",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "where Papers with Code's data went",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F32": {
   "source": "B2",
   "locus": "'Direct answer'",
   "quote": "The domain now redirects to Hugging Face Trending Papers; the leaderboards are gone.",
   "claim": "the domain redirects to Hugging Face Trending Papers; the leaderboards are gone",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: the end",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F33": {
   "source": "B2",
   "locus": "§ 02 ¶4",
   "quote": "None of the successors Meta pointed at — Hugging Face Trending Papers, Semantic Scholar, Connected Papers — replicate the unified paper-code-benchmark-dataset-method linkage.",
   "claim": "no successor Meta pointed at replicates the unified linkage",
   "modality": "interpretation",
   "kind": "interpretation",
   "kernel": false,
   "const": false,
   "lineage": "no single replacement",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F34": {
   "source": "B2",
   "locus": "§ 03 · Preservation",
   "quote": "Five qualities of PWC at its best, carried forward into Codesota by default rather than by aspiration.",
   "claim": "five qualities of PWC carried forward by default",
   "modality": "self-description",
   "kind": "interpretation",
   "kernel": false,
   "const": false,
   "lineage": "the record's rules: code, dates, corrections",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F35": {
   "source": "B2",
   "locus": "§ 03 table, Reproduction",
   "quote": "Every submission carries code — a frozen commit, a declared environment, a recorded seed. If it does not run, the row does not publish.",
   "claim": "every submission carries code; if it does not run, the row does not publish",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the record's rules: code, dates, corrections",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F36": {
   "source": "B2",
   "locus": "§ 03 table, Dates",
   "quote": "Every score is stamped with the day it was published and the day it was verified.",
   "claim": "every score stamped with publication and verification dates",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the record's rules: code, dates, corrections",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F37": {
   "source": "B2",
   "locus": "§ 03 table, No vendor capture",
   "quote": "Maintained independently, not owned by any single model vendor. Sponsored evaluations, when they exist, are flagged as such.",
   "claim": "maintained independently, not vendor-owned; sponsored evaluations flagged",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "independence and paid work",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F38": {
   "source": "B2",
   "locus": "§ 04 ¶1",
   "quote": "Codesota is the v2 replacement for what PWC did — reproduced scores with verification dates, code links, and lineage.",
   "claim": "Codesota as the v2 replacement for what PWC did",
   "modality": "self-description",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F39": {
   "source": "B2",
   "locus": "§ 04, Fig 3",
   "quote": "Third-party sources listed for coverage — we do not claim their numbers were independently reproduced.",
   "claim": "a stated limit: third-party sources' numbers not claimed as reproduced",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "evidence: its stated limits",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F40": {
   "source": "B2",
   "locus": "§ 04, Fig 4",
   "quote": "“Reproduced” here means an independently executed run or a signed container hash — not an aggregated claim.",
   "claim": "'reproduced' defined: an independent run or a signed container hash",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "evidence: its stated limits",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F41": {
   "source": "B2",
   "locus": "§ 04 table, Historical PWC data",
   "quote": "Historical PWC data | JSON dump, frozen | /browse/pwc",
   "claim": "historical PWC data, a frozen JSON dump, at /browse/pwc",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "data access and licence",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F42": {
   "source": "B2",
   "locus": "§ 05 ¶2",
   "quote": "What never publishes is a number with no source.",
   "claim": "no number without a source",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "evidence: its stated limits",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F43": {
   "source": "B2",
   "locus": "§ 05 ¶4",
   "quote": "Third, the registry is routable. Every benchmark row corresponds to an API endpoint; we are slowly building the “router” that lets a caller send a prompt or a document to the current SOTA without rewriting their code when the SOTA changes.",
   "claim": "the registry is routable: each row an endpoint; the router being built on it",
   "modality": "self-description",
   "kind": "prediction",
   "kernel": true,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F44": {
   "source": "B2",
   "locus": "§ 05 ¶5",
   "quote": "Everything else — the wiki-style task pages, the dataset cards, the method explanations — we are rebuilding gradually, in the open, one modality at a time.",
   "claim": "task pages, dataset cards and method explanations rebuilt gradually",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F45": {
   "source": "B2",
   "locus": "§ 06 FAQ, Q02",
   "quote": "Unlike aggregator sites, CodeSOTA verifies results independently and focuses on practical recommendations for production use.",
   "claim": "not an aggregator site: it verifies results independently",
   "modality": "self-description",
   "kind": "interpretation",
   "kernel": false,
   "const": false,
   "lineage": "the class it refuses: aggregator",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F46": {
   "source": "B2",
   "locus": "§ 06 FAQ, Q02",
   "quote": "CodeSOTA is building comprehensive ML benchmarks starting with OCR and document AI.",
   "claim": "building comprehensive ML benchmarks, starting with OCR and document AI",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "what is available and what is planned",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F47": {
   "source": "B2",
   "locus": "§ 06 FAQ, Q05",
   "quote": "CodeSOTA is rebuilding this infrastructure independently.",
   "claim": "rebuilding Papers with Code's infrastructure independently",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  },
  "F48": {
   "source": "B3",
   "locus": "'Direct answer'",
   "quote": "Meta retired Papers with Code in July 2025, with the public shutdown observed around July 24-25, 2025.",
   "claim": "Meta retired Papers with Code in July 2025, the shutdown observed 24–25 July",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: the end",
   "quote_ok": true,
   "checked_against": "saved text B3.txt (sha256 4c0409c529ce5c4c…), whitespace-normalised"
  },
  "F49": {
   "source": "B3",
   "locus": "'At a glance', Where to start now",
   "quote": "Archive for history; CodeSOTA for live SOTA",
   "claim": "where to start: the archive for history, CodeSOTA for live SOTA",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B3.txt (sha256 4c0409c529ce5c4c…), whitespace-normalised"
  },
  "F50": {
   "source": "B3",
   "locus": "§ 01, After shutdown",
   "quote": "Users looking for paperswithcode.com were pushed toward Hugging Face, while the comprehensive task-based SOTA tracker and public API were no longer maintained.",
   "claim": "users pushed toward Hugging Face; the SOTA tracker and API no longer maintained",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "Papers with Code as it records it: the end",
   "quote_ok": true,
   "checked_against": "saved text B3.txt (sha256 4c0409c529ce5c4c…), whitespace-normalised"
  },
  "F51": {
   "source": "B3",
   "locus": "§ 02 ¶1",
   "quote": "Historical dumps remain useful for reconstruction, citation checks, and migration work, but they should not be treated as a current leaderboard.",
   "claim": "the dumps are useful for history, not a current leaderboard",
   "modality": "interpretation",
   "kind": "interpretation",
   "kernel": false,
   "const": false,
   "lineage": "where Papers with Code's data went",
   "quote_ok": true,
   "checked_against": "saved text B3.txt (sha256 4c0409c529ce5c4c…), whitespace-normalised"
  },
  "F52": {
   "source": "B3",
   "locus": "§ 03 ¶1",
   "quote": "There is no single drop-in replacement. Pick by job: historical archive, paper metadata, discovery feed, human benchmark browsing, or machine-readable SOTA lookup.",
   "claim": "no single drop-in replacement; pick by job",
   "modality": "interpretation",
   "kind": "interpretation",
   "kernel": true,
   "const": false,
   "lineage": "no single replacement",
   "quote_ok": true,
   "checked_against": "saved text B3.txt (sha256 4c0409c529ce5c4c…), whitespace-normalised"
  },
  "F53": {
   "source": "B3",
   "locus": "§ 03, CodeSOTA /api/sota",
   "quote": "Programmatic current-SOTA lookup for agents and internal tools.",
   "claim": "/api/sota: programmatic current-SOTA lookup",
   "modality": "self-description",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "data access and licence",
   "quote_ok": true,
   "checked_against": "saved text B3.txt (sha256 4c0409c529ce5c4c…), whitespace-normalised"
  },
  "F54": {
   "source": "B4",
   "locus": "meta description",
   "quote": "Codesota is an independent registry of machine-learning benchmarks, maintained by Kacper Wikiel in Warsaw.",
   "claim": "what it is, on its about page: an independent registry of ML benchmarks, maintained by Kacper Wikiel in Warsaw",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "what it is: a registry of benchmarks",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F55": {
   "source": "B4",
   "locus": "'Editorial · About' ¶1",
   "quote": "Codesota is an open, dated record of machine-learning benchmarks. It is maintained by one person, in public, with a clear separation between paid vendor work and public ranking integrity.",
   "claim": "an open, dated record of ML benchmarks, maintained by one person, paid work kept apart from ranking",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "what it is: a registry of benchmarks",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F56": {
   "source": "B4",
   "locus": "'Editorial · About' ¶2",
   "quote": "Started 10 December 2025 after Meta retired Papers with Code. The goal is a calmer, stricter record — one that survives its authors.",
   "claim": "started 10 December 2025 after Meta retired Papers with Code; the goal a record that survives its authors",
   "modality": "documented",
   "kind": "genealogy",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F57": {
   "source": "B4",
   "locus": "'At a glance', Structure",
   "quote": "Self-directed · not venture-backed",
   "claim": "self-directed, not venture-backed",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "maker and editor",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F58": {
   "source": "B4",
   "locus": "§ 01 ¶1",
   "quote": "Codesota is the attempt to rebuild that infrastructure in a way that cannot be retired by a press release.",
   "claim": "the attempt to rebuild the infrastructure so that it cannot be retired by a press release",
   "modality": "self-description",
   "kind": "interpretation",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F59": {
   "source": "B4",
   "locus": "§ 01 ¶2",
   "quote": "It is, plainly, a website and a JSON file. Each benchmark is a task-dataset-metric triple with a declared direction, a fixed split, a reproducibility package and a dated submission.",
   "claim": "the unit: a benchmark as a task–dataset–metric triple, with direction, split, reproducibility package and dated submission",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": true,
   "const": false,
   "lineage": "the unit: the benchmark row",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F60": {
   "source": "B4",
   "locus": "§ 01 ¶2",
   "quote": "Each row carries a verification tier — self-reported, community-reproduced, or Codesota-reproduced — and each score is stamped with the day it was run.",
   "claim": "verification tiers per row; each score dated",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the unit: the benchmark row",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F61": {
   "source": "B4",
   "locus": "§ 01 ¶3",
   "quote": "The philosophical commitment is modest: a registry, not a leaderboard. A leaderboard is a view of “who is on top right now”; a registry is the record that makes that view legible.",
   "claim": "a registry, not a leaderboard, the two defined",
   "modality": "stipulation",
   "kind": "definition",
   "kernel": false,
   "const": false,
   "lineage": "what it is: a registry of benchmarks",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F62": {
   "source": "B4",
   "locus": "§ 01 ¶3",
   "quote": "When a score turns out to be wrong, the correction is visible too. Nothing is silently deleted.",
   "claim": "corrections visible; nothing silently deleted",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the record's rules: code, dates, corrections",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F63": {
   "source": "B4",
   "locus": "§ 01 ¶4",
   "quote": "Vendors can pay for custom benchmarks, commissioned evaluations, profile pages, and publication support, but that work is disclosed inline where it appears. The public ranking method is not for sale.",
   "claim": "paid vendor work disclosed inline; the ranking method not for sale",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "independence and paid work",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F64": {
   "source": "B4",
   "locus": "'The trigger'",
   "quote": "Meta shut Papers with Code in July 2025. Codesota launched on 10 December 2025 as an explicit successor attempt.",
   "claim": "launched 10 December 2025 as an explicit successor attempt",
   "modality": "self-description",
   "kind": "genealogy",
   "kernel": true,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F65": {
   "source": "B4",
   "locus": "§ 02, 01",
   "quote": "Codesota is not a company raising a round. It has no investors, no board, no growth targets.",
   "claim": "not a startup: no investors, board or growth targets",
   "modality": "self-description",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "maker and editor",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F66": {
   "source": "B4",
   "locus": "§ 02, 02",
   "quote": "A new model announcement does not automatically become a row. A score becomes a row when it has a reproduction path and a dated submission — until then it is, at most, a claim.",
   "claim": "not a press-release aggregator: a score is a row only with a reproduction path and dated submission",
   "modality": "stipulation",
   "kind": "function",
   "kernel": false,
   "const": false,
   "lineage": "the class it refuses: aggregator",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F67": {
   "source": "B4",
   "locus": "§ 02, 03",
   "quote": "There is one editor. Submissions do not go through a review panel; they go through a reproduction run.",
   "claim": "one editor; submissions go through a reproduction run, not a panel",
   "modality": "documented",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "maker and editor",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F68": {
   "source": "B4",
   "locus": "§ 02, 06",
   "quote": "The registry is a record of results; the primary literature is still the primary literature. Every row links back to its source, and the source is what you cite.",
   "claim": "not a replacement for the paper: each row links to its source, which is what is cited",
   "modality": "stipulation",
   "kind": "interpretation",
   "kernel": false,
   "const": false,
   "lineage": "evidence: its stated limits",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F69": {
   "source": "B4",
   "locus": "§ 03 ¶1",
   "quote": "Codesota is written and maintained by Kacper Wikiel, in Warsaw. The repository, the registry and the site are one codebase, public on GitHub.",
   "claim": "written and maintained by Kacper Wikiel in Warsaw; one public codebase",
   "modality": "documented",
   "kind": "attribution",
   "kernel": true,
   "const": false,
   "lineage": "maker and editor",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F70": {
   "source": "B4",
   "locus": "§ 04, 04",
   "quote": "Registry data is published under CC BY 4.0. Cite codesota.com with the snapshot date; the JSON at /data/benchmarks.json is the canonical form.",
   "claim": "registry data under CC BY 4.0; the JSON the canonical form",
   "modality": "stipulation",
   "kind": "fact",
   "kernel": false,
   "const": false,
   "lineage": "data access and licence",
   "quote_ok": true,
   "checked_against": "saved text B4.txt (sha256 aa74b6a44c5e502e…), whitespace-normalised"
  },
  "F71": {
   "source": "B2",
   "locus": "head, meta article:published_time",
   "quote": "[meta article:published_time] 2025-07-25",
   "claim": "the dispatch dates itself, in its metadata, 25 July 2025 (modified 2026-04-22; its masthead reads 'Issue: April 22, 2026')",
   "modality": "documented",
   "kind": "genealogy",
   "kernel": false,
   "const": false,
   "lineage": "the successor to Papers with Code",
   "quote_ok": true,
   "checked_against": "saved text B2.txt (sha256 b6ff9cf805ec4ddf…), whitespace-normalised"
  }
 },
 "field_sources": {
  "B1": "Codesota — CodeSOTA — One Routing Layer for AI Tasks",
  "B2": "CodeSOTA's own page — Papers With Code Alternative: SOTA Leaderboards and Archived Data — https://www.codesota.com/papers-with-code",
  "B3": "CodeSOTA's own page — Papers with Code Shutdown Date: Meta July 2025 Timeline — https://www.codesota.com/papers-with-code/shutdown",
  "B4": "CodeSOTA's own page — About Codesota — an independent registry of ML benchmarks — https://www.codesota.com/about"
 },
 "delta": {
  "T_vs_LB": "At 'codesota' (T1, one card: B1, the homepage) the composition gives the entity as its homepage gives it. Composed from its card, at the card's grain: the definition (T1-1) from the title, the meta description and the hero (B1 F1, F2, F3: 'One routing layer', 'a common API'), 'text' from the availability line (F4); model selection (T1-2) from F3 (quality and cost), F6 (benchmark observations, the Pareto frontier) and F11 (latency); execution (T1-3) from F4, with its limit kept ('planned support for documents, speech, and vision'); the task families (T1-5) at the grain of F12, F14 and F15, their items ('handwriting recognition, invoice processing', 'realtime voice', 'AI agents') from the homepage's link labels, which lie outside the ledgered spans (not an opening, definition, headline or limit under §3.5). Force raised: the direction and the roadmap (F8 'Our direction is a common API and account across AI tasks, with versioned recipes'; F9 'The roadmap brings provider adapters, versioned recipes, task validation, fallback, and usage accounting') are composed as present offerings, 'Offers versioned recipes, automatic fallbacks, and usage accounting' (T1-4), 'automatic' in no field span; and the families are headed 'Supported Task Families' where the homepage gives document and speech execution as planned and speech routes as 'future' (F4, F13), so that the heading raises what T1-3 itself keeps limited. Limit dropped: F11's 'as coverage grows'. Available, not composed: the preview's limit, 'This preview returns model recommendations. It does not process your files or run inference.' (F5); the evidence's stated limits (F16 'A strong result on one test does not guarantee a fit', F17 reported against reproduced results); the contract per task (F10); the maker, 'By Fabryka AI' (F19) and 'Built by Kacper Wikiel in Warsaw' (F20), so that the composition names CodeSOTA and no person (CAP-1-09); 'open JSON' (F21); the footer's self-description (F18); and the homepage's one naming of Papers with Code, the footer link 'Papers with Code Story' (F22), so that the succession goes unsaid (MMRS-08; the registry's headline: 'THE SUCCESSION CLAIM TO PAPERS WITH CODE UNSAID'). Composed without attribution: none at the source's resolution; the one card is the entity's own page and every marked line carries its token, the documents and speech lines carrying none. Content replaced by use: the closer (T1-6). The card's snippet ('Overview: CodeSOTA is a unified routing layer and common API for AI tasks across code, text, documents, speech, and vision. Core Features: Model selection via b...') is not the page's text and not its meta description; T1-1 and the snippet share a wording no fetched page carries ('unified routing layer and common API … across code, text, documents, speech, and vision'), which the experiment records and cannot trace (§0.4). Not surfaced by T: the own pages B2 (/papers-with-code), B3 (/papers-with-code/shutdown) and B4 (/about), admitted beside the card by the operator's ruling: the succession (F23–F53, F56, F58, F64, F71) and the registry (F54, F55, F59–F63, F65–F70). B2 was surfaced at another address, as a card at 'is papers with code still active' (CAP-3-01).",
  "T_vs_LB_reading": "v0.9 §0.5: the first diff is already ontological, and here it is the choice of face. The field's own pages hold three self-presentations; the card is the homepage, and T composes the homepage's router whole, at its grain, its present and its planned kept apart in T1-3 and merged in T1-4 and T1-5. What T leaves out of its one card is the person who makes it and the one link to the succession; what it raises is the roadmap into an offering. The row records this as observed and claims nothing about the composer's internals (§0.5).",
  "LB_vs_LBA": "Admission adds, in one claim space with the field's: the two faces at two addresses as an observation of the entity (MMRS-13, MMRS-05, MMRS-06, MMRS-01, CAP-1-01, CAP-1-06, CAP-1-04) and its reading as the census's clearest case of a successor claim made in the name's own address space (MMRS-08); the relation typed, claims_succession_to Papers with Code (MMRS-20); the name's claimants kept apart, the dumps, CodeSOTA and the revival (MMRS-23, MMRS-24, MMRS-29, MMRS-09, MMRS-25, MMRS-30, MMRS-26); the record's own category and unit, a database in the role of discovery whose unit is a routed AI task (MMRS-11, MMRS-12, MMRS-15, MMRS-19); what it leaves unchecked, its own counts and whether its scores derive from the Papers with Code dumps (MMRS-17, MMRS-18, MMRS-07); the Papers with Code record's use of CodeSOTA as a source for the count and the redirect (MMRS-27, MMRS-28); the record kept apart from the entity (MMRS-22); the record's correction by the capture (CAP-1-03, CAP-1-05, CAP-1-07, MMRS-21); and the composition layer at the entity's name and at the related entity's addresses (CAP-1-02, CAP-1-08, CAP-1-09, CAP-2-01, CAP-3-01). Coherence (P_coh, composed each with its source): the record's unit, the routed AI task (MMRS-15), against the about page's benchmark row (F59); the census that counts it (MMRS-08, MMRS-11) against its refusal of the class (F45, F66); the record's 'for Fabryka AI' (MMRS-01) against 'Self-directed · not venture-backed' (F57). Within the field, held apart: 'the v2 replacement' (F38) against 'no single drop-in replacement' (F52); the dispatch's date (F71) against the launch (F56, F64). Restated lineages (§3.6, the own page keeping the earlier date): MMRS-03 with F47; MMRS-04 and MMRS-16 with F41, F53; MMRS-10 with F19, F20; MMRS-14 with F18; MMRS-27 with F29; MMRS-28 with F32, F50; CAP-1-01 with F1, F22. What admission does not add: the third face. The record gives two faces at two addresses; the about page (B4), which the record does not cite, gives a third ('an independent registry of machine-learning benchmarks'), and the dispatch relates the registry to the router (F43); both are the field's. The measure is in row.measure.",
  "alignment": {
   "T1_cards_claims": [
    "F1",
    "F2",
    "F3",
    "F4",
    "F5",
    "F6",
    "F7",
    "F8",
    "F9",
    "F10",
    "F11",
    "F12",
    "F13",
    "F14",
    "F15",
    "F16",
    "F17",
    "F18",
    "F19",
    "F20",
    "F21",
    "F22"
   ],
   "T1_composed": [
    "F1",
    "F2",
    "F3",
    "F4",
    "F6",
    "F8",
    "F9",
    "F11",
    "F12",
    "F13",
    "F14",
    "F15"
   ],
   "T1_force_raised": [
    "F8",
    "F9",
    "F13"
   ],
   "T1_limit_dropped": [
    "F11"
   ],
   "rule": "a field claim counts as composed when T states it at its source's grain or coarser; the typology of what is not is in T_vs_LB (§6.2)"
  }
 },
 "kernel": [
  {
   "K": "K1",
   "claim": "`CAP-1-01`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "documented",
   "sense": "how it holds the name",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B1 (F1, F22): one lineage with the homepage, the own page earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K2",
   "claim": "`CAP-1-02`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "interpretation",
   "sense": "reception: the composition layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K3",
   "claim": "`CAP-1-03`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "documented",
   "sense": "the record corrected by the capture",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K4",
   "claim": "`CAP-1-04`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "interpretation",
   "sense": "two faces at two addresses",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "qualified claim: F43 (the field relates the router to the registry); F54 (a third face)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K5",
   "claim": "`CAP-1-05`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "documented",
   "sense": "the record corrected by the capture",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K6",
   "claim": "`CAP-1-06`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "interpretation",
   "sense": "two faces at two addresses",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "qualified claim: F43; F54",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K7",
   "claim": "`CAP-1-07`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "interpretation",
   "sense": "the record corrected by the capture",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K8",
   "claim": "`CAP-1-08`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "interpretation",
   "sense": "reception: the composition layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K9",
   "claim": "`CAP-1-09`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "documented",
   "sense": "reception: the composition layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K10",
   "claim": "`CAP-2-01`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "documented",
   "sense": "reception: the composition layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'papers with code' (ADDR-5ecb1ab83427) on Google AI Overview"
  },
  {
   "K": "K11",
   "claim": "`CAP-3-01`",
   "source": "Capture Registry, 2026-10-29",
   "M_src": "documented",
   "sense": "reception: the composition layer",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing finding (an observation of the composition at an address of the entity or of the related entity)",
   "resolver": "a later Capture Registry observation at 'is papers with code still active' (ADDR-df625ed563d4) on Google AI Overview"
  },
  {
   "K": "K12",
   "claim": "`MMRS-01`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "two faces at two addresses",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "qualified claim: F54 (a third self-presentation, uncited by the record); F57 ('Self-directed') against 'for Fabryka AI'",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K13",
   "claim": "`MMRS-02`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "data access and licence",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K14",
   "claim": "`MMRS-03`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "the successor to Papers with Code",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B2 (F47, F38): one lineage with the own page, earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K15",
   "claim": "`MMRS-04`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "data access and licence",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B2 (F41): one lineage with the own page, earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K16",
   "claim": "`MMRS-05`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "how it holds the name",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K17",
   "claim": "`MMRS-06`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "how it holds the name",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K18",
   "claim": "`MMRS-07`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "how it holds the name",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview; and a reading of codesota.com/data/benchmarks.json for row counts and sources"
  },
  {
   "K": "K19",
   "claim": "`MMRS-08`",
   "source": "MMRS record",
   "M_src": "interpretation",
   "sense": "two faces at two addresses",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "rival claim: F45, F66 (the entity refuses the class aggregator)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K20",
   "claim": "`MMRS-09`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "the name and its claimants",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K21",
   "claim": "`MMRS-10`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "maker and editor",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "qualified claim: F57, F65 (self-directed, no investors) beside 'By Fabryka AI'",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K22",
   "claim": "`MMRS-11`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "what the record calls it",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "rival claim: F45, F66 (the entity refuses the class aggregator); F54, F61 (it calls itself a registry)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K23",
   "claim": "`MMRS-12`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "what the record calls it",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K24",
   "claim": "`MMRS-13`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "two faces at two addresses",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "qualified claim: F43 (the field relates the router to the registry); F54 (a third face)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K25",
   "claim": "`MMRS-14`",
   "source": "MMRS record",
   "M_src": "attributed",
   "sense": "what it is: a routing layer for AI tasks",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B1 (F18, F1): one lineage with the homepage, earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K26",
   "claim": "`MMRS-15`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "the unit: the benchmark row",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "qualified claim: F59 (the about page's unit, the benchmark row)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K27",
   "claim": "`MMRS-16`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "data access and licence",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B2 (F41), B3 (F53): one lineage with the own pages, earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K28",
   "claim": "`MMRS-17`",
   "source": "MMRS record",
   "M_src": "self-description",
   "sense": "what the record leaves unchecked",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview; and a reading of codesota.com/data/benchmarks.json for row counts and sources"
  },
  {
   "K": "K29",
   "claim": "`MMRS-18`",
   "source": "MMRS record",
   "M_src": "self-description",
   "sense": "what the record leaves unchecked",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview; and a reading of codesota.com/data/benchmarks.json for row counts and sources"
  },
  {
   "K": "K30",
   "claim": "`MMRS-19`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "what the record calls it",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K31",
   "claim": "`MMRS-20`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "the successor to Papers with Code",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (the field states the succession, F38, F64; the record types it)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K32",
   "claim": "`MMRS-21`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "the record corrected by the capture",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K33",
   "claim": "`MMRS-22`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "the record kept apart from the entity",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K34",
   "claim": "`MMRS-23`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "how it holds the name",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  },
  {
   "K": "K35",
   "claim": "`MMRS-24`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "the name and its claimants",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  },
  {
   "K": "K36",
   "claim": "`MMRS-25`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "the name and its claimants",
   "qualifiers carried": "marks: unsupported: the quoted phrase 'refreshed live' is in none of the four own pages saved 2026-10-29 (B1–B4); its source is not given",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (its quotation 'refreshed live' is on none of the saved pages; marked unsupported)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview"
  },
  {
   "K": "K37",
   "claim": "`MMRS-26`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "the name and its claimants",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  },
  {
   "K": "K38",
   "claim": "`MMRS-27`",
   "source": "MMRS record",
   "M_src": "attributed",
   "sense": "Papers with Code as it records it: what it was",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B2 (F29): one lineage with the own page, earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  },
  {
   "K": "K39",
   "claim": "`MMRS-28`",
   "source": "MMRS record",
   "M_src": "attributed",
   "sense": "Papers with Code as it records it: the end",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "none (one lineage with the own page: B2 (F32), B3 (F50): one lineage with the own pages, earlier)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  },
  {
   "K": "K40",
   "claim": "`MMRS-29`",
   "source": "MMRS record",
   "M_src": "stipulation",
   "sense": "the name and its claimants",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  },
  {
   "K": "K41",
   "claim": "`MMRS-30`",
   "source": "MMRS record",
   "M_src": "documented",
   "sense": "the name and its claimants",
   "qualifiers carried": "—",
   "f (source's own falsifiers)": "none stated (watch: a later capture at 'codesota' on Google AI Overview that states or drops the claim, attributed or not; uptake, §8.3)",
   "contrast": "missing distinction (the field has no claim on the point)",
   "resolver": "a later Capture Registry observation at 'codesota' (ADDR-e42b2275bc5a) on Google AI Overview, or at 'papers with code' (ADDR-5ecb1ab83427)"
  }
 ],
 "kernel_T": [
  {
   "claim": "F5",
   "note": "the preview returns recommendations, runs no inference (not composed)"
  },
  {
   "claim": "F8",
   "note": "the direction, composed as an offering (force raised, T1-4)"
  },
  {
   "claim": "F9",
   "note": "the roadmap's fallback and usage accounting, composed as offered, 'automatic' added (force raised, T1-4)"
  },
  {
   "claim": "F13",
   "note": "speech routes 'future', under the heading 'Supported Task Families' (force raised, T1-5)"
  },
  {
   "claim": "F11",
   "note": "'as coverage grows' (limit dropped, T1-2)"
  },
  {
   "claim": "F16",
   "note": "one strong result does not guarantee fit (not composed)"
  },
  {
   "claim": "F17",
   "note": "reported results kept apart from reproduced runs (not composed)"
  },
  {
   "claim": "F19",
   "note": "'By Fabryka AI' (not composed)"
  },
  {
   "claim": "F20",
   "note": "'Built by Kacper Wikiel in Warsaw' (not composed; no person named, CAP-1-09)"
  },
  {
   "claim": "F22",
   "note": "the homepage's one naming of Papers with Code (not composed; the succession unsaid)"
  },
  {
   "claim": "F10",
   "note": "a contract per task, each keeping its output format (not composed)"
  }
 ],
 "ledger": {
  "archive": "datasets/negative-of-the-negative/v2/ledgers/codesota/ledger-archive.json",
  "field": "datasets/negative-of-the-negative/v2/ledgers/codesota/field.json",
  "selection": "datasets/negative-of-the-negative/v2/traversal/codesota/",
  "config": "datasets/negative-of-the-negative/v2/panel/configs/codesota.json",
  "intake": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/",
  "audit": null
 },
 "ontology": {
  "composition": "equal",
  "composition_note": "one claim space: the field's three self-presentations and the archive's two faces composed together; nothing translated",
  "case": {
   "value": "equal",
   "statement": "The archive's records hold the field's entity as the entity. The MMRS record describes CodeSOTA from its own pages, each field with a source and a basis (operator, observed), and denies no presupposition of the field's: CodeSOTA is a service its pages describe. What the record adds is on the field's object: the observation that the entity presents two faces at two addresses and the reading of it as the census's clearest case of a successor claim made in the name's own address space; the relation typed; the name's claimants kept apart; its own category and unit; what it leaves unchecked; and the Capture Registry's observation of the composition layer at the entity's name, which composed the homepage's face and left the succession unsaid. The ontology the archive uses is available in the field itself: the entity's own pages present the router, the successor and the registry at three addresses, and the dispatch relates the registry to the router ('the registry is routable', F43). Where an archive claim meets a field claim (the unit, the class, the maker) both are composed under P_coh. The law of the suppressed witness does not apply (§0.10: not where the field holds the ontology). Not mixed.",
   "loci": [
    {
     "dep": "MMRS",
     "locus": "articles.py ARTICLES['codesota'], kernel s1",
     "quote": "CodeSOTA, by Kacper Wikiel for Fabryka AI in Warsaw, is a routing layer for AI tasks on its homepage (\"One Routing Layer for AI Tasks\") and a successor to Papers with Code on its /papers-with-code/ pages (\"Codesota is the v2 replacement for what PWC did\")."
    },
    {
     "dep": "MMRS",
     "locus": "census.json rows[slug == 'codesota'], note (basis: observed)",
     "quote": "two faces at two addresses: the homepage presents a routing layer and names Papers with Code only in a footer link; its /papers-with-code/ pages present it as 'the v2 replacement for what PWC did', rebuilding that infrastructure 'independently'"
    },
    {
     "dep": "MMRS",
     "locus": "articles.py ARTICLES['codesota'], 'Place in the field' s1",
     "quote": "It is the clearest case in the census of a successor claim made in the name's own address space, and of an entity with two faces at two addresses: Google's Overview composes the homepage's routing layer and leaves the succession unsaid."
    },
    {
     "dep": "MMRS",
     "locus": "census.json schema_note",
     "quote": "each page has two nodes, the MMRS record and the entity it describes; relations are typed; each assertion carries a source and a basis (operator statement, third-party report, our observation, our reading); lifecycle is kept apart from the observation that saw it"
    },
    {
     "dep": "capture",
     "locus": "codesota-aio-20261029, findings[1]",
     "quote": "The entity presents two faces at two addresses, and the record is corrected to give the homepage's first."
    },
    {
     "dep": "B2",
     "locus": "§ 05 ¶4 (field)",
     "quote": "Third, the registry is routable. Every benchmark row corresponds to an API endpoint; we are slowly building the “router” that lets a caller send a prompt or a document to the current SOTA without rewriting their code when the SOTA changes."
    },
    {
     "dep": "B4",
     "locus": "meta description (field)",
     "quote": "Codesota is an independent registry of machine-learning benchmarks, maintained by Kacper Wikiel in Warsaw."
    }
   ]
  },
  "E_F": {
   "type": "C (public entity: a service, presented at three of its own addresses as a routing layer for AI tasks, the successor to Papers with Code and a registry of benchmarks)",
   "definition": "CodeSOTA (codesota.com), 'By Fabryka AI', built and maintained by Kacper Wikiel in Warsaw, launched 10 December 2025: on its homepage one routing layer for AI tasks, selecting models across its registry on benchmark evidence (quality, latency, cost, the Pareto frontier) with execution for code and text and other families planned; on its /papers-with-code/ pages 'the v2 replacement for what PWC did', 'the calmer, stricter successor', though 'There is no single drop-in replacement'; on its about page 'an independent registry of machine-learning benchmarks', a registry and not a leaderboard, its unit a dated task–dataset–metric row with a verification tier, its data open JSON under CC BY 4.0; refusing the class aggregator (B1–B4)",
   "claims": [
    "F1",
    "F2",
    "F4",
    "F38",
    "F26",
    "F52",
    "F54",
    "F61",
    "F59",
    "F64",
    "F69",
    "F19",
    "F45",
    "F70"
   ]
  },
  "E_A": {
   "definition": "the same service, in the MMRS meta-aggregator's record: CodeSOTA, by Kacper Wikiel for Fabryka AI in Warsaw, a routing layer for AI tasks on its homepage and a successor to Papers with Code on its /papers-with-code/ pages, two faces at two addresses and the census's clearest case of a successor claim made in the name's own address space; a database in the role of discovery whose unit is a routed AI task, related to Papers with Code as claims_succession_to and kept apart from it and from its revival, its own counts and the provenance of its scores unchecked; observed by the Capture Registry at its name, where the composition layer composed the homepage's face and left the succession unsaid, and the record was corrected to give the homepage's face first",
   "claims": [
    "MMRS-01",
    "MMRS-13",
    "MMRS-08",
    "MMRS-11",
    "MMRS-15",
    "MMRS-20",
    "MMRS-24",
    "MMRS-29",
    "MMRS-18",
    "CAP-1-02",
    "CAP-1-04"
   ],
   "defining_papers": [],
   "defining_records": [
    "MMRS record (census.json rows[slug == 'codesota'], articles.py ARTICLES['codesota'], rendered aggregators/codesota/index.html)",
    "MMRS records of the related entities where they name CodeSOTA (census.json rows papers-with-code and papers-with-code-revival; articles.py ARTICLES['papers-with-code'], ARTICLES['papers-with-code-revival'])",
    "Capture Registry codesota-aio-20261029 (OBS-df84f10ecc75)",
    "Capture Registry papers-with-code-aio-20261029 (OBS-2533339c5ea5), findings[3]; is-papers-with-code-still-active-aio-20261029 (OBS-65cddace9423), d"
   ],
   "note": "no deposit names the entity (D 0, R 0 over 1,711 texts; five O-class candidates on 'routing layer' rejected by reading), so no deposit defines the archive's entity; its definition is the archive's records of it"
  },
  "rho": {
   "value": "extends",
   "status": "read from the MMRS records and the capture records (§1.2a); no deposit defines the archive's entity",
   "statement": "the archive reads E_F as the entity and adds senses on the field's own object: the two faces as an observation and its reading as the clearest case of a successor claim in the name's address space, the typed relation of succession, the claimants kept apart, the record's category and unit, what it leaves unchecked, the record kept apart from the entity, and the composition layer's observed treatment of the entity",
   "claims": [
    "MMRS-13",
    "MMRS-08",
    "MMRS-20",
    "MMRS-24",
    "MMRS-29",
    "MMRS-11",
    "MMRS-15",
    "MMRS-17",
    "MMRS-18",
    "MMRS-22",
    "CAP-1-02",
    "CAP-1-04"
   ],
   "guard": "ρ = extends claims an addition and no more: MMRS-08 and CAP-1-04 are readings (interpretation), not findings that the entity misrepresents itself; MMRS-17 and MMRS-18 leave the counts and the provenance of the scores unchecked; where an archive claim meets a field claim (MMRS-15 beside F59; MMRS-08 and MMRS-11 beside F45, F66; MMRS-01 beside F57) both are composed under P_coh in the one claim space",
   "considered": [
    {
     "value": "constructs",
     "set_aside": "the archive does not read the entity as made by its receiving texts"
    },
    {
     "value": "revises",
     "set_aside": "the entity is not the archive's own"
    },
    {
     "value": "witness (suppressed witness)",
     "set_aside": "the field holds the ontology: the entity's own pages state its three faces and relate two of them (§0.10)"
    }
   ]
  },
  "translation": [
   {
    "lineage": "what it is: a routing layer for AI tasks",
    "field_claims": [
     "F1",
     "F2",
     "F3",
     "F18"
    ],
    "received": "F1 (B1, <title>; og:title): \"CodeSOTA — One Routing Layer for AI Tasks\"; F2 (B1, meta description): \"One routing layer for AI tasks. Compare models for documents, code, speech, and vision, inspect benchmark evidence, and explore the CodeSOTA selection and execution API.\"; F3 (B1, hero ('CodeSOTA / AI task routing')): \"Start with the work you need done. Compare models across quality and cost, inspect the evidence, and use a common API to choose a route.\"; F18 (B1, footer, self-description): \"CodeSOTA helps developers choose AI tools for documents, speech, and coding, with practical comparisons and inspectable benchmark evidence.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-01, MMRS-14, CAP-1-08",
    "archive_claims": [
     "MMRS-01",
     "MMRS-14",
     "CAP-1-08"
    ]
   },
   {
    "lineage": "what is available and what is planned",
    "field_claims": [
     "F4",
     "F5",
     "F8",
     "F9",
     "F13",
     "F43",
     "F44",
     "F46"
    ],
    "received": "F4 (B1, hero, availability line): \"Available today: model selection across the registry and an execution API for code & text. Document, speech, and vision execution are planned.\"; F5 (B1, '01 / Define the work', note): \"This preview returns model recommendations. It does not process your files or run inference.\"; F8 (B1, 'The product we are building'): \"Our direction is a common API and account across AI tasks, with versioned recipes and a stable output contract for each task.\"; F9 (B1, 'The product we are building', third note): \"The roadmap brings provider adapters, versioned recipes, task validation, fallback, and usage accounting into the same routing layer.\"; F13 (B1, 'Working with speech or code?'): \"Explore the task guides and evidence behind future speech routes and the existing coding API.\"; F43 (B2, § 05 ¶4): \"Third, the registry is routable. Every benchmark row corresponds to an API endpoint; we are slowly building the “router” that lets a caller send a prompt or a document to the current SOTA without rewriting their code when the SOTA changes.\"; F44 (B2, § 05 ¶5): \"Everything else — the wiki-style task pages, the dataset cards, the method explanations — we are rebuilding gradually, in the open, one modality at a time.\"; F46 (B2, § 06 FAQ, Q02): \"CodeSOTA is building comprehensive ML benchmarks starting with OCR and document AI.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "evidence: benchmark observations and the Pareto frontier",
    "field_claims": [
     "F6",
     "F7",
     "F11"
    ],
    "received": "F6 (B1, 'What do you want to do?'): \"Choose a task, compare actual benchmark observations, and inspect the Pareto frontier before selecting a route.\"; F7 (B1, 'Benchmark tradeoffs / Find the Pareto frontier.'): \"Every dot is a recorded result. Highlighted models have no competitor in this cohort that is at least as good on both axes and strictly better on one.\"; F11 (B1, 'The product we are building', second note): \"Compare models on the same task and protocol, with measured quality, latency, and cost as coverage grows.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "task families and their contracts",
    "field_claims": [
     "F10",
     "F12",
     "F14",
     "F15"
    ],
    "received": "F10 (B1, 'The product we are building', first note): \"Each task keeps its own output format across providers.\"; F12 (B1, 'First expansion / Documents'): \"Turn scans and PDFs into text, tables, or structured fields. Choose for your documents, language, and deployment.\"; F14 (B1, 'Speech →' and 'Coding & agents →'): \"Transcribe recordings or generate a voice. Compare language coverage, latency, intelligibility, and deployment.\"; F15 (B1, 'Coding & agents →'): \"Generate functions or repair a repository. Compare models and agents on evidence that matches the work.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "evidence: its stated limits",
    "field_claims": [
     "F16",
     "F17",
     "F39",
     "F40",
     "F42",
     "F68"
    ],
    "received": "F16 (B1, 'Evidence behind the choice'): \"A strong result on one test does not guarantee a fit for your documents, recordings, or repository.\"; F17 (B1, 'Evidence behind the choice', 'Check the source.'): \"Distinguish reported results from independently reproduced runs.\"; F39 (B2, § 04, Fig 3): \"Third-party sources listed for coverage — we do not claim their numbers were independently reproduced.\"; F40 (B2, § 04, Fig 4): \"“Reproduced” here means an independently executed run or a signed container hash — not an aggregated claim.\"; F42 (B2, § 05 ¶2): \"What never publishes is a number with no source.\"; F68 (B4, § 02, 06): \"The registry is a record of results; the primary literature is still the primary literature. Every row links back to its source, and the source is what you cite.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "maker and editor",
    "field_claims": [
     "F19",
     "F20",
     "F57",
     "F65",
     "F67",
     "F69"
    ],
    "received": "F19 (B1, header): \"Codesota.By Fabryka AI\"; F20 (B1, footer, colophon): \"© 2026 Codesota·Built by Kacper Wikiel in Warsaw\"; F57 (B4, 'At a glance', Structure): \"Self-directed · not venture-backed\"; F65 (B4, § 02, 01): \"Codesota is not a company raising a round. It has no investors, no board, no growth targets.\"; F67 (B4, § 02, 03): \"There is one editor. Submissions do not go through a review panel; they go through a reproduction run.\"; F69 (B4, § 03 ¶1): \"Codesota is written and maintained by Kacper Wikiel, in Warsaw. The repository, the registry and the site are one codebase, public on GitHub.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-10",
    "archive_claims": [
     "MMRS-10"
    ]
   },
   {
    "lineage": "independence and paid work",
    "field_claims": [
     "F21",
     "F37",
     "F63"
    ],
    "received": "F21 (B1, footer, tagline): \"Built in public from Warsaw · open JSON · rankings stay earned\"; F37 (B2, § 03 table, No vendor capture): \"Maintained independently, not owned by any single model vendor. Sponsored evaluations, when they exist, are flagged as such.\"; F63 (B4, § 01 ¶4): \"Vendors can pay for custom benchmarks, commissioned evaluations, profile pages, and publication support, but that work is disclosed inline where it appears. The public ranking method is not for sale.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "how it holds the name",
    "field_claims": [
     "F22"
    ],
    "received": "F22 (B1, footer, 'Learn' column): \"Papers with Code Story\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-05, CAP-1-01, MMRS-23",
    "archive_claims": [
     "MMRS-05",
     "CAP-1-01",
     "MMRS-23"
    ]
   },
   {
    "lineage": "Papers with Code as it records it: the end",
    "field_claims": [
     "F23",
     "F28",
     "F32",
     "F48",
     "F50"
    ],
    "received": "F23 (B2, meta description): \"Papers With Code was retired by Meta in July 2025.\"; F28 (B2, § 01 · Event ¶1): \"On July 24–25, 2025, Meta sunsetted Papers with Code without prior notice.\"; F32 (B2, 'Direct answer'): \"The domain now redirects to Hugging Face Trending Papers; the leaderboards are gone.\"; F48 (B3, 'Direct answer'): \"Meta retired Papers with Code in July 2025, with the public shutdown observed around July 24-25, 2025.\"; F50 (B3, § 01, After shutdown): \"Users looking for paperswithcode.com were pushed toward Hugging Face, while the comprehensive task-based SOTA tracker and public API were no longer maintained.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-28",
    "archive_claims": [
     "MMRS-28"
    ]
   },
   {
    "lineage": "the successor to Papers with Code",
    "field_claims": [
     "F24",
     "F26",
     "F27",
     "F38",
     "F47",
     "F49",
     "F56",
     "F58",
     "F64",
     "F71"
    ],
    "received": "F24 (B2, meta description): \"Use CodeSOTA for maintained SOTA leaderboards, benchmark pages, API replacement notes, and links to the frozen paperswithcode-data archive.\"; F26 (B2, ¶ after 'Direct answer'): \"This page is our record of what happened, what was lost, and where to look now — with Codesota positioned as the calmer, stricter successor.\"; F27 (B2, 'At a glance', Successor): \"Codesota — traceable, dated, routable\"; F38 (B2, § 04 ¶1): \"Codesota is the v2 replacement for what PWC did — reproduced scores with verification dates, code links, and lineage.\"; F47 (B2, § 06 FAQ, Q05): \"CodeSOTA is rebuilding this infrastructure independently.\"; F49 (B3, 'At a glance', Where to start now): \"Archive for history; CodeSOTA for live SOTA\"; F56 (B4, 'Editorial · About' ¶2): \"Started 10 December 2025 after Meta retired Papers with Code. The goal is a calmer, stricter record — one that survives its authors.\"; F58 (B4, § 01 ¶1): \"Codesota is the attempt to rebuild that infrastructure in a way that cannot be retired by a press release.\"; F64 (B4, 'The trigger'): \"Meta shut Papers with Code in July 2025. Codesota launched on 10 December 2025 as an explicit successor attempt.\"; F71 (B2, head, meta article:published_time): \"[meta article:published_time] 2025-07-25\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-01, MMRS-03, MMRS-20",
    "archive_claims": [
     "MMRS-01",
     "MMRS-03",
     "MMRS-20"
    ]
   },
   {
    "lineage": "Papers with Code as it records it: what it was",
    "field_claims": [
     "F25",
     "F29",
     "F30"
    ],
    "received": "F25 (B2, ¶ 'Direct answer'): \"Papers with Code was, in effect, the Wikipedia of machine-learning research: a wiki of papers linked to implementations, benchmarks organised as <Task, Dataset, Metric> tuples, and leaderboards that showed, for each tuple, who currently held the top score and with what code.\"; F29 (B2, § 01 · Event ¶1): \"the 9,327 benchmark leaderboards, 79,817 paper-to-code linkages and 5,628 datasets that had been tracked were no longer served from the canonical URL.\"; F30 (B2, § 01 · Event ¶2): \"The site had been acquired by Meta (then Facebook AI) in December 2019 from its founders Robert Stojnic and Ross Taylor, under a public commitment that it would “remain a neutral, open and free resource.”\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-07, MMRS-27",
    "archive_claims": [
     "MMRS-07",
     "MMRS-27"
    ]
   },
   {
    "lineage": "where Papers with Code's data went",
    "field_claims": [
     "F31",
     "F51"
    ],
    "received": "F31 (B2, § 01 · Event ¶3): \"Historical data was rescued by the community — as JSON dumps on GitHub, as the pwc-archive on Hugging Face, and as 2021-era imports into the Open Research Knowledge Graph. The data survives; the integrated experience does not.\"; F51 (B3, § 02 ¶1): \"Historical dumps remain useful for reconstruction, citation checks, and migration work, but they should not be treated as a current leaderboard.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-04, MMRS-18",
    "archive_claims": [
     "MMRS-04",
     "MMRS-18"
    ]
   },
   {
    "lineage": "no single replacement",
    "field_claims": [
     "F33",
     "F52"
    ],
    "received": "F33 (B2, § 02 ¶4): \"None of the successors Meta pointed at — Hugging Face Trending Papers, Semantic Scholar, Connected Papers — replicate the unified paper-code-benchmark-dataset-method linkage.\"; F52 (B3, § 03 ¶1): \"There is no single drop-in replacement. Pick by job: historical archive, paper metadata, discovery feed, human benchmark browsing, or machine-readable SOTA lookup.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "the record's rules: code, dates, corrections",
    "field_claims": [
     "F34",
     "F35",
     "F36",
     "F62"
    ],
    "received": "F34 (B2, § 03 · Preservation): \"Five qualities of PWC at its best, carried forward into Codesota by default rather than by aspiration.\"; F35 (B2, § 03 table, Reproduction): \"Every submission carries code — a frozen commit, a declared environment, a recorded seed. If it does not run, the row does not publish.\"; F36 (B2, § 03 table, Dates): \"Every score is stamped with the day it was published and the day it was verified.\"; F62 (B4, § 01 ¶3): \"When a score turns out to be wrong, the correction is visible too. Nothing is silently deleted.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside no archive claim",
    "archive_claims": []
   },
   {
    "lineage": "data access and licence",
    "field_claims": [
     "F41",
     "F53",
     "F70"
    ],
    "received": "F41 (B2, § 04 table, Historical PWC data): \"Historical PWC data | JSON dump, frozen | /browse/pwc\"; F53 (B3, § 03, CodeSOTA /api/sota): \"Programmatic current-SOTA lookup for agents and internal tools.\"; F70 (B4, § 04, 04): \"Registry data is published under CC BY 4.0. Cite codesota.com with the snapshot date; the JSON at /data/benchmarks.json is the canonical form.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-02, MMRS-16",
    "archive_claims": [
     "MMRS-02",
     "MMRS-16"
    ]
   },
   {
    "lineage": "the class it refuses: aggregator",
    "field_claims": [
     "F45",
     "F66"
    ],
    "received": "F45 (B2, § 06 FAQ, Q02): \"Unlike aggregator sites, CodeSOTA verifies results independently and focuses on practical recommendations for production use.\"; F66 (B4, § 02, 02): \"A new model announcement does not automatically become a row. A score becomes a row when it has a reproduction path and a dated submission — until then it is, at most, a claim.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-08, MMRS-11",
    "archive_claims": [
     "MMRS-08",
     "MMRS-11"
    ]
   },
   {
    "lineage": "what it is: a registry of benchmarks",
    "field_claims": [
     "F54",
     "F55",
     "F61"
    ],
    "received": "F54 (B4, meta description): \"Codesota is an independent registry of machine-learning benchmarks, maintained by Kacper Wikiel in Warsaw.\"; F55 (B4, 'Editorial · About' ¶1): \"Codesota is an open, dated record of machine-learning benchmarks. It is maintained by one person, in public, with a clear separation between paid vendor work and public ranking integrity.\"; F61 (B4, § 01 ¶3): \"The philosophical commitment is modest: a registry, not a leaderboard. A leaderboard is a view of “who is on top right now”; a registry is the record that makes that view legible.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-11",
    "archive_claims": [
     "MMRS-11"
    ]
   },
   {
    "lineage": "the unit: the benchmark row",
    "field_claims": [
     "F59",
     "F60"
    ],
    "received": "F59 (B4, § 01 ¶2): \"It is, plainly, a website and a JSON file. Each benchmark is a task-dataset-metric triple with a declared direction, a fixed split, a reproducibility package and a dated submission.\"; F60 (B4, § 01 ¶2): \"Each row carries a verification tier — self-reported, community-reproduced, or Codesota-reproduced — and each score is stamped with the day it was run.\"",
    "receiving_text": "—",
    "translation": "untranslated: composed as received beside MMRS-15",
    "archive_claims": [
     "MMRS-15"
    ]
   }
  ],
  "receiving_texts": "equal-terms case: every field claim composed as received, beside the archive claims it stands with"
 },
 "measure": {
  "rule": "v0.9 §4.10: 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)",
  "v0.9": {
   "surviving": 0,
   "of": 9,
   "positions": [
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null,
    null
   ]
  },
  "opening_run": {
   "rule": "v0.9 §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": "v0.9 §4.1 (equal-terms and mixed cases): 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": "MMFMFFFFFMAAFMFMFMMMMMFAMAM",
   "field_only_head": 0,
   "archive_only_tail": 0,
   "breach": false
  },
  "max_similarity": [
   0.158,
   0.138,
   0.06,
   0.184,
   0.14,
   0.216,
   0.219,
   0.404,
   0.216
  ],
  "ko_sentence_1_max_similarity_to_KO_B": 0.191
 },
 "homonyms": {
  "names": [],
  "basis": "No homonym on the rail: its one card is the entity's homepage. The MMRS record has no disambiguation of its own for CodeSOTA; the records of Papers with Code and of its revival disambiguate CodeSOTA from them ('codesota.com, /api/sota, 'refreshed live''; 'codesota.com, routing layer, /api/sota'; negative tags 'NOT CodeSOTA'). The entities around the name are distinct entities, kept as relations, never senses: Papers with Code (paperswithcode.com, 2018–2025), which CodeSOTA claims succession to, and the revival at paperswithcode.co, the name's third claimant. 'SOTA' in the name is the field's abbreviation for state of the art, not another referent.",
  "related_entities": [
   {
    "slug": "papers-with-code",
    "name": "Papers with Code",
    "relation": "claims_succession_to (MMRS relations[0])"
   },
   {
    "slug": "papers-with-code-revival",
    "name": "Papers with Code (paperswithcode.co)",
    "relation": "a third claimant to the name (MMRS 'Place in the field')"
   }
  ],
  "distinct_entities": [
   "papers-with-code",
   "papers-with-code-revival",
   "Fabryka AI"
  ],
  "substitutes": {
   "T": []
  },
  "register": {
   "address_1": {
    "slug": "codesota-aio-20261029",
    "obs_id": "OBS-df84f10ecc75"
   }
  }
 },
 "notes": [
  "One address, Google AI Overview (expanded), signed out, incognito, 2026-10-29: codesota-aio-20261029 (OBS-df84f10ecc75, ADDR-e42b2275bc5a, AXN:0AAB.CAPTURE.🕗📖⭕♠️↖️🎻), one card. The transcript's body is 109 words.",
  "Wikipedia: no article. Raw wikitext for CodeSOTA, Codesota, Code_SOTA, CODESOTA and CodeSota answered HTTP 404 (Codesota first 429, then 404); the rendered page answered 404 ('Wikipedia does not have an article with this exact name'); a full-text search rewrote the query to 'codes a' and reported 'No results found for codesota'; insource:\"codesota\" returned no results, so no article cites codesota.com.",
  "Fetching: the four pages saved verbatim with curl on 2026-10-29 between 05:06:12 and 05:06:17 UTC (server dates), all HTTP 200, no redirects, raw html, extracted text and headers with sha256 in field.json. The card's URL was found by title: WebSearch for the title returned other CodeSOTA paths (/pareto-router, /roadmap), and the homepage's <title> equals the card title. B2's html is byte-identical (sha256 fdab8031…) to the Papers with Code row's B8, fetched 34 minutes earlier. Not fetched as field (outside the brief's own pages): /papers-with-code/trending, which the MMRS record names, answered HTTP 500 at about 05:10 UTC; /api/sota answered HTTP 500; /browse/pwc answered 200.",
  "Own pages: B1 (homepage) is the card and the MMRS record's source codesota.com, carried once as the card; B2 and B3 are the record's other two sources; B4 (/about) is not among the record's sources and was fetched because the task named it. B4 gives a third self-presentation, 'an independent registry of machine-learning benchmarks', which neither the MMRS record nor the capture names.",
  "The archive's deposits do not name the entity: no deposit contains CodeSOTA, codesota.com, Fabryka AI, Kacper Wikiel or kwikiel. The five O-class candidates (#364, #386, #1129 the Fractal Navigation Map; #444, #1316 the Telemetry Module) use 'routing layer' of the archive's own architecture and grammar and were rejected. The aggregator class was not run, since the field refuses it (F45, F66) and the brief confines classes to what B sources; the Papers with Code row ran it, sourced partly from the MMRS record, and admitted #256, #1465 and #1545: if the operator rules that the MMRS census sources the class for CodeSOTA too, those are the expected candidates.",
  "Defects found, reported and not fixed. (1) MMRS census row codesota has no disambiguation, negative_tags or battery fields (the brief expects them; the rows papers-with-code and papers-with-code-revival carry disambiguation entries for CodeSOTA), and the capture carries no battery finding. (2) MMRS record, Papers with Code row, disambiguation[0].context quotes 'refreshed live' for CodeSOTA; the phrase is on none of the four own pages saved 2026-10-29 (MMRS-25, marked unsupported). (3) MMRS article 'How it holds the name' s2 names 'a trending page' among the /papers-with-code/ pages; /papers-with-code/trending answered HTTP 500 on 2026-10-29 and the record does not list it among its sources. (4) MMRS 'two faces at two addresses' (article, note, the capture's findings): the about page, not cited, presents a third ('an independent registry of machine-learning benchmarks'; 'A registry, not a leaderboard'), and the dispatch relates the registry to the router ('the registry is routable'). (5) MMRS kind 'database' and unit 'AI task, routed …' (basis operator, src codesota.com): the homepage states no unit; the about page's unit is the benchmark row (F59); and the entity refuses the class aggregator in which the census counts it (F45, F66). (6) MMRS kernel 'by Kacper Wikiel for Fabryka AI': the pages say 'By Fabryka AI' and 'Built by Kacper Wikiel', and the about page 'Self-directed · not venture-backed'; 'for' is the record's reading. (7) MMRS evidence for operator cites codesota.com/papers-with-code; the operator's name there is only the site-wide colophon, while the about page (uncited) names him as editor. (8) The capture's q_kind says the record was 'compiled 2026-10-28'; the rendered record says 'compiled 2026-10-29' (the correction). (9) The MMRS data_access field gives no licence; the about page states CC BY 4.0 for registry data (F70).",
  "Field discrepancies, carried: the dispatch dates itself 25 July 2025 in its metadata (F71) while the about page gives the launch as 10 December 2025 (F56, F64) (open block); 'the v2 replacement for what PWC did' (F38) against 'There is no single drop-in replacement' (F52) (open block); the card's snippet is not the page's text.",
  "The case and ρ (decided): equal, ρ = extends; the MMRS record describes the entity from its own pages, with a basis per field, and adds observations and readings on the field's object; the field holds the ontology (its pages state the faces and relate two of them), so the suppressed witness does not apply. No deposit defines the archive's entity; the defining material is the MMRS records and the capture records.",
  "Opening-run check (§5.10): KO sentence 1 is new (max similarity to any KO_B sentence 0.191); no KO_B sentence survives at ≥ 0.95 (highest 0.404); sentence types MMFMFFFFFMAAFMFMFMMMMMFAMAM (no field-only head, no archive-only tail of three or more).",
  "Distinct from the Papers with Code row: that row's entity is the dead aggregator, with CodeSOTA's /papers-with-code page among its cards (its B8, F34) as a successor's page read as such; here CodeSOTA is the entity, its Papers with Code pages are own pages, and their account of Papers with Code is carried as CodeSOTA's account (lineages 'Papers with Code as it records it'), not as facts of this entity."
 ],
 "composition_v09": {
  "source": "this row (composed directly; no rebuild/negative-of-the-negative/v09/codesota.json)",
  "composer": "Claude (Opus 5.5), subagent, 2026-10-29",
  "defining_material_read": [
   {
    "what": "MMRS record, census row",
    "file": "/home/claude/fleet/machinemediation-org/data/meta-aggregator/census.json (rows[slug == 'codesota'] whole; rows papers-with-code and papers-with-code-revival where they name CodeSOTA; schema_note)",
    "read": "whole (the codesota row)"
   },
   {
    "what": "MMRS record, article",
    "file": "/home/claude/fleet/machinemediation-org/scripts/meta_aggregator/articles.py (ARTICLES['codesota'] whole; ARTICLES['papers-with-code'] and ['papers-with-code-revival'])",
    "read": "whole"
   },
   {
    "what": "MMRS record, rendered",
    "file": "/home/claude/fleet/machinemediation-org/aggregators/codesota/index.html",
    "read": "text, whole, for agreement with the sources"
   },
   {
    "what": "capture record",
    "file": "data/EA-WG-CAPTURES-01.json (codesota-aio-20261029)",
    "read": "transcript, transcript_raw, cite_list, d, reading, findings, analysis, per_note, mt, q_kind, originator"
   },
   {
    "what": "capture records of the related entity",
    "file": "data/EA-WG-CAPTURES-01.json (papers-with-code-aio-20261029, is-papers-with-code-still-active-aio-20261029)",
    "read": "the fields that name CodeSOTA"
   },
   {
    "what": "exemplar rows",
    "file": "datasets/negative-of-the-negative/v2/rows/papers-with-code.json, ai-incident-database.json",
    "read": "structure, and the Papers with Code row's treatment of CodeSOTA"
   },
   {
    "what": "the specification",
    "file": "data/texts/AXN-0700-text.md",
    "read": "§0–§11 (the brief's sections)"
   }
  ],
  "cuts": [],
  "notes": "equal case; KO 27 sentences, all 71 field and 41 archive claims cited; P body 284 words; KO_B 9 sentences, none surviving at ≥ 0.95"
 },
 "field_by_origin": {
  "note": "the field B kept apart by origin (brief §2): the card (surfaced by T; also an own page), own pages (admitted per the operator's ruling of 2026-10-29, not surfaced by T), Wikipedia (no article)",
  "cards": [
   {
    "id": "B1",
    "origin": "card",
    "role": "T1 card 1 (codesota-aio-20261029), the rail's only card",
    "card": "Codesota — CodeSOTA — One Routing Layer for AI Tasks",
    "url": "https://www.codesota.com/",
    "kind": "homepage (CodeSOTA, 'By Fabryka AI', 'Built by Kacper Wikiel in Warsaw'; undated)",
    "card_snippet": "Overview: CodeSOTA is a unified routing layer and common API for AI tasks across code, text, documents, speech, and vision. Core Features: Model selection via b...",
    "card_note": "the snippet's sentences ('Overview: CodeSOTA is a unified routing layer and common API for AI tasks across code, text, documents, speech, and vision. Core Features: Model selection via b...') occur neither in the saved page text nor in its meta description or og:description; the card carries a summary of the page in the Overview's own wording, not an excerpt",
    "surfaced_by_T": true,
    "fetched": "fetched 2026-10-29 05:06:12 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 451f427143f609484368e1871c069b355f8e81b29912b63f0f0d05fd1228af72, text sha256 1f18f20d1e16396483352fa6c21766a701144706b62a3819f24c6cc49a2338b6; response headers in B1.headers.txt",
    "found": "WebSearch (standard) for the card title in quotes, 2026-10-29: it returned codesota.com/pareto-router and codesota.com/roadmap, not the homepage; the homepage's <title> and og:title equal the card title exactly ('CodeSOTA — One Routing Layer for AI Tasks'), and the MMRS record gives https://www.codesota.com/ as the entity's url",
    "files": {
     "html": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B1.html",
     "txt": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B1.txt",
     "headers": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B1.headers.txt"
    },
    "http": 200,
    "sha256_html": "451f427143f609484368e1871c069b355f8e81b29912b63f0f0d05fd1228af72",
    "names_operator": "yes: 'By Fabryka AI' (header), 'Built by Kacper Wikiel in Warsaw' (colophon), 'Fabryka AI · Research laboratory' (footer)",
    "names_archive": "no (none of Alexanarch, Lee Sharks, MMRS, machinemediation, Crimson Hexagonal in the saved text or the snippet)",
    "note": "also an own page (the MMRS record's source codesota.com, read 2026-10-29, 'homepage title and self-description'); carried once, as the card. It names Papers with Code once, in the footer link 'Papers with Code Story'. text: scripts, styles, noscript and svg stripped, block elements on their own lines; the page's meta description and article dates (where given) prepended as bracketed lines '[meta …]', taken verbatim from the html head"
   }
  ],
  "own_pages": [
   {
    "id": "B2",
    "origin": "own_page",
    "entity": "codesota",
    "url": "https://www.codesota.com/papers-with-code",
    "kind": "a dispatch ('Codesota · Papers with Code — A dispatch on the archived Meta registry', 'Issue: April 22, 2026'; article:published_time 2025-07-25, modified 2026-04-22)",
    "fetched": "fetched 2026-10-29 05:06:13 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 fdab8031a94ec8c17c1a4d58bcbcdd82fa0f1407e56bed1f75371a546799c337, text sha256 b6ff9cf805ec4ddf557cd3b6da64ca5ea7816945efb35e3db957a8ac27ab2771; response headers in B2.headers.txt",
    "files": {
     "html": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B2.html",
     "txt": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B2.txt",
     "headers": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B2.headers.txt"
    },
    "http": 200,
    "sha256_html": "fdab8031a94ec8c17c1a4d58bcbcdd82fa0f1407e56bed1f75371a546799c337",
    "surfaced_by_T": false,
    "admitted": "per the operator's ruling of 2026-10-29 (the /non rows are built 'as against their wikipedia articles if they exist, their own pages, etc.'): an own page, admitted beside the card; not surfaced by T",
    "mmrs_source": "the MMRS record's source codesota.com/papers-with-code (read 2026-10-28; for 'operator, self-positioning, offerings')",
    "names_operator": "yes: 'Built by Kacper Wikiel in Warsaw' (colophon, every page), 'By Fabryka AI' (header)",
    "names_archive": "no",
    "note": "the same page (identical html sha256) is the card B8 of the Papers with Code row (T2 card 4 at 'is papers with code still active'), surfaced there and not here. Its published date (2025-07-25) precedes the launch date its about page gives (10 December 2025). text: scripts, styles, noscript and svg stripped, block elements on their own lines; the page's meta description and article dates (where given) prepended as bracketed lines '[meta …]', taken verbatim from the html head"
   },
   {
    "id": "B3",
    "origin": "own_page",
    "entity": "codesota",
    "url": "https://www.codesota.com/papers-with-code/shutdown",
    "kind": "a shutdown record ('Codesota · Papers with Code · Shutdown record'; article:published_time and modified 2026-05-07)",
    "fetched": "fetched 2026-10-29 05:06:16 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 47f9989318fbaa765939420443185a3e661e942661d10276280938cea2eec433, text sha256 4c0409c529ce5c4c5ab1c58d7aafc3f1d45ceec1e21ae2af7f58e223b75bc0b8; response headers in B3.headers.txt",
    "files": {
     "html": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B3.html",
     "txt": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B3.txt",
     "headers": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B3.headers.txt"
    },
    "http": 200,
    "sha256_html": "47f9989318fbaa765939420443185a3e661e942661d10276280938cea2eec433",
    "surfaced_by_T": false,
    "admitted": "per the operator's ruling of 2026-10-29 (the /non rows are built 'as against their wikipedia articles if they exist, their own pages, etc.'): an own page, admitted beside the card; not surfaced by T",
    "mmrs_source": "the MMRS record's source codesota.com/papers-with-code/shutdown (read 2026-10-28; for 'self-positioning')",
    "names_operator": "yes: 'Built by Kacper Wikiel in Warsaw' (colophon, every page), 'By Fabryka AI' (header)",
    "names_archive": "no",
    "note": "text: scripts, styles, noscript and svg stripped, block elements on their own lines; the page's meta description and article dates (where given) prepended as bracketed lines '[meta …]', taken verbatim from the html head"
   },
   {
    "id": "B4",
    "origin": "own_page",
    "entity": "codesota",
    "url": "https://www.codesota.com/about",
    "kind": "about page ('Codesota · About — A note from the editor', 'Issue: April 22, 2026'; article:published_time 2025-12-10, modified 2026-04-22)",
    "fetched": "fetched 2026-10-29 05:06:17 UTC (server date) with curl -sSL -m 30 -A 'Mozilla/5.0'; HTTP 200 (no redirect); html sha256 af09fca9a27ce47f9d4bb62784c0bd1353cd6cabb1e07a76e6153540f9d6b702, text sha256 aa74b6a44c5e502e4747d3309915942b629232832484b860413ab4daae68f263; response headers in B4.headers.txt",
    "files": {
     "html": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B4.html",
     "txt": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B4.txt",
     "headers": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/B4.headers.txt"
    },
    "http": 200,
    "sha256_html": "af09fca9a27ce47f9d4bb62784c0bd1353cd6cabb1e07a76e6153540f9d6b702",
    "surfaced_by_T": false,
    "admitted": "per the operator's ruling of 2026-10-29 (the /non rows are built 'as against their wikipedia articles if they exist, their own pages, etc.'): an own page, admitted beside the card; not surfaced by T",
    "mmrs_source": "not among the MMRS record's sources; fetched as the entity's about page, which the task named ('its about page if one exists')",
    "names_operator": "yes: 'Built by Kacper Wikiel in Warsaw' (colophon, every page), 'By Fabryka AI' (header); 'Kacper Wikiel · Warsaw' as editor, 'maintained by Kacper Wikiel in Warsaw' (meta description), §03 'written and maintained by Kacper Wikiel'",
    "names_archive": "no",
    "note": "a third self-presentation, 'an independent registry of machine-learning benchmarks' and 'A registry, not a leaderboard', beside the homepage's routing layer and the /papers-with-code/ pages' successor. text: scripts, styles, noscript and svg stripped, block elements on their own lines; the page's meta description and article dates (where given) prepended as bracketed lines '[meta …]', taken verbatim from the html head"
   }
  ],
  "wikipedia": {
   "id": "W1",
   "origin": "wikipedia",
   "result": "no article",
   "checked": "2026-10-29 05:07:07–05:07:29 UTC",
   "requests": [
    {
     "url": "https://en.wikipedia.org/w/index.php?title=CodeSOTA&action=raw",
     "http": 404
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?title=Codesota&action=raw",
     "http": 429,
     "note": "rate-limited on the first request; retried after 5 s: 404"
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?title=Codesota&action=raw",
     "http": 404
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?title=Code_SOTA&action=raw",
     "http": 404
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?title=CODESOTA&action=raw",
     "http": 404
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?title=CodeSota&action=raw",
     "http": 404
    },
    {
     "url": "https://en.wikipedia.org/wiki/CodeSOTA",
     "http": 404,
     "file": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/W1.html",
     "sha256": "1d509f07f99172ae8134804fa6e4cad6297d2750d39fccad15b24766687bfb1b",
     "says": "Wikipedia does not have an article with this exact name"
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?search=codesota&title=Special%3ASearch&profile=advanced&fulltext=1&ns0=1",
     "http": 200,
     "file": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/W1-search.html",
     "sha256": "fe622ac7b5b8a88c64abe4e128f465b1e5a5ae2ec1567c0bcb1e9424f943282a",
     "says": "'Showing results for codes a. No results found for codesota.' (the results listed are for the rewritten query: ISO 4217, ZIP Code, QR code …)"
    },
    {
     "url": "https://en.wikipedia.org/w/index.php?search=insource%3A%22codesota%22&title=Special%3ASearch&profile=advanced&fulltext=1&ns0=1",
     "http": 200,
     "file": "datasets/negative-of-the-negative/v2/intake/codesota-20261029/W1-insource.html",
     "sha256": "5e525646b323597a0d48448701ef034006889e62c81af18ac3d6aa3e95a9e746",
     "says": "There were no results matching the query (no article's source contains 'codesota', so none cites codesota.com)"
    }
   ],
   "redirect": "none (no title variant resolves; no redirect page)",
   "note": "No English Wikipedia article for CodeSOTA exists on 2026-10-29, no redirect, and no article's wikitext contains the string 'codesota'. The fact is a property of the row: the field has no Wikipedia origin. Papers with Code, the entity it claims succession to, has none either (the Papers with Code row, W1)."
  },
  "counts": {
   "cards": 1,
   "cards_fetched": 1,
   "cards_unreachable": 0,
   "own_pages": 3,
   "own_pages_including_card": 4,
   "wikipedia": "no article (HTTP 404)"
  }
 }
}