A v1.1 adversarial technical referee report by Lee Sharks, with Nobel Glas and Talos Morrow, reviewing Giovanni Battista Caria’s *AI Bleeding: Semantic Exhaustion via Out-of-Distribution Linguistic Payload*.
The report argues that the reviewed paper does not establish a novel linguistic resource-exhaustion attack. It identifies five claimed defects in the paper’s own evidence: a negative and nonsignificant…
A Wound Gauge case report by Lee Sharks documenting a source-reported five-round Google AI Overview battery on June 13, 2026. The query concerned the *Self-Audit Module for Public Summarizers*, a named five-metric provenance instrument. According to the report, the initial output replaced the module with a generic corporate-style checklist, omitted its authorship and DOI, then evaluated itself with the…
A v1.1 synthesis paper by Talos Morrow and Nobel Glas, edited by Lee Sharks, connecting model collapse, AI-generated-code security, and software monoculture through the proposed variable **solution-space diversity**.
The paper argues that collapse in code may remain invisible to correctness benchmarks because generated programs can continue to compile and pass tests while converging on a narrower set of…
A thesis-level MPAI disambiguation packet by Lee Sharks separating the **Josephus Thesis** from the Piso Hypothesis and from broader Christ-myth theories.
The Josephus Thesis is presented as an available heteronymic reading of New Testament authorship: John, Paul, Peter, James, Luke, Mark, and Matthew may be read as differentiated author-functions within a Josephan literary system formed under Roman capture. The…
A v1.0 protocol specification by Lee Sharks defining the **SPXI Self-Audit Block**, a compact provenance and meaning-preservation structure to be embedded in documents intended for machine retrieval and summarization.
The block declares author, identifier, institution, designation, canonical key terms, a provenance kernel, audit targets, a non-erasure condition, and a canary test. It is designed to make…
This document is a living work plan for extracting and governing the Crimson Hexagonal Archive’s coined vocabulary. Its initial phase templates say “not started,” but the later progress table and session log record substantial execution. It should therefore be read as a plan whose headings preserve the original design while its ledger records the state reached during the same work cycle.
The planned workflow has…
This draft paper audits 1,817 DOI strings associated with the Crimson Hexagonal Archive and distinguishes three layers commonly collapsed under “persistent identifier”: the character string, URL resolution, and publicly recoverable identifying metadata.
The reconciled result reports 946 DOI records returning public DataCite metadata and 871 returning HTTP 404 from the DataCite public API on June 22, 2026. The…
The June 22 pre-cleanup scan applies the first Surface Weather Station method to a nine-object battery. It records hand-coded visibility, anchor, integrity, and lift values; a dashboard; object-level readings; and predictions for how the surface should respond to link repointing and sovereign-site cleanup.
Its values are historical and retrieval-environment-specific. Later versions of the method prohibit several…
This representative scan covers eight objects, marks four as not executed, links a canonical JSON record, and compares its results with parallel substrate readings. It confirms successor-anchor lag and ghost survival while locating sharp cross-system divergence around PER, Writable Retrieval Basin, Revelation First, and Alexanarch.
Its principal value is methodological rather than the particular dashboard score.…
Version 1.1.1 hardens the Surface Weather Station after technical review, the first federated scan, and substrate-identification failures. It is the first version to cleanly separate search retrieval from scoring, lock expected figures and query strings, preserve raw evidence independently of annotation, and account for backend properties that alter visibility.
The specification intentionally holds weights,…
A documentary structural argument built around CERN’s public technology-transfer narrative and Berners-Lee’s public critique of the Web’s institutional evolution. It analyzes five anchor points, maps the critique to the governance-competence framework, and proposes originator-audit as an MMRS methodological precedent.
The inference from Web governance history to CERN’s competence to operate scholarly…
A close reading of four CERN ODP portal pages captured during an active data-subject request. The work moves through three interpretive iterations—from textual contradiction, to architectural degradation, to subject-composing gesture—and fixes the third as the deposited analysis.
It introduces subject-composing operative metadata and pairs the portal’s longitudinal tunnel with the laterally distributed…
Consolidated chronology of the AI Overview Capture Registry's final week from three independent source classes — registry metadata (DataCite), platform timestamps (TikTok), contemporaneous public statements (GitHub #2606) — with deponent attestation marked and subordinated. Anchors: concept DOI registered 2026-06-13T22:48:57Z; counter at 309 views / 666 downloads captured and platform-timestamped 2026-06-15…
Measured answer to 'is 1,000 downloads in days a lot by Zenodo's standards': all 1,791 datasets published in the registry's publication window (2026-06-14..16), n=200 sampled via the public API, stats read at day ≈19. Cohort median 7 downloads; p95 116; p99 804 (version-level). Corrected anchors (v2 erratum, platform-timestamp-driven, superseded text preserved): 666 cumulative at 2026-06-15 16:42 EDT; 1,000+ at…
Empirical research instrument for measuring platform-driven scholarly-record erasure. Companion deposit to EA-DATAHUB-01 v1.0 (AXN:0420, GOVERNANCE): where DATAHUB is the strategic plan, EROSION is the instrument. Phase 1 (running): the Zenodo Removal Census, preserving Zenodo's monthly deleted-records CSV (1,309,351 rows in the 2026-06-07 snapshot; Zenodo retains only three most recent snapshots) with…
An empirical audit of Zenodo's public deletion enumeration across a 33-day window (2026-06-07 to 2026-07-10, exports containing 1,309,351 and 1,322,007 identifier entries respectively), testing the two-part asymmetry predicted by EA-PROVENANCE-DEBT-01 §3 at the outcome level.
Three findings are established at full evidentiary strength. First, programmed bibliographic suppression: Zenodo's public deletion…
The anti-suppression paper, deposit of the developmental version: full eight-section text with appendices A–F, produced through five-substrate Assembly Chorus consultation (Rounds 1–4 plus final checks), carrying its own falsification battery (sixteen tests), the container model under the content-agnostic ruling (core/sidecar determined fully by the operative mint; remint and stratification; roles not kinds), the…
Source-engaged supersession of the developmental-version deposit AXN:0457 (#1094), composed under the standing rule (2026-07-18, decisions register): engagement with sources is composition-primary in formal academic works. §§II (Historical Precedents) and III (Technical Precedents) — the sections that survey — re-composed reading the scholarship that documents each case (Doucette, Komaromi, Johns, Benkler, Levy,…
Reference capture #1 for a proposed new MMRS capture category, provisionally named AGGREGATOR_RENDER (naming pending MANUS ruling; alternates: TAXONOMIC_CAPTURE, NORMALIZATION_CAPTURE, EXTRACTOR_RENDER, RECEPTION_TEMPLATE). The SciLynk display of THE SHARKS ARK: EA-ARK-01-ARCHON v3.0 (screencapped 2026-07-18) fits no existing Capture Registry category: SciLynk performs aggregator normalization, not generation.…
The known-bad corpus offered to GoogleCloudPlatform/knowledge-catalog issue #207 and taken up for conformance fixtures: 111 cases across 16 classes drawn from the 2026-06-19 termination event and its aftermath, each expressed in the thread's ratified axes (validity / presence / edges; presence sparse with exactly removed and never_landed as named states; reason objects with not_disclosed first-class; the…
Successor to deposit #1455. Ten pre-registered tests, each with its prediction committed before its run, clearing the seven-item correction ledger of v0.2 and generating four further tests in response to results. Two predictions confirmed, two partially confirmed, six missed, one registered panel not run, one line closed with a specific handoff, and seventeen deviations between registration and execution…
Synthesis of four independent Assembly reviews of the measurement-stack documents. Eight points of unanimity, including that the integration note governs, that Transit Rent is the only new measure, that DS-6 stays fixed at six, and that the self-correction chain is the credibility asset rather than a weakness. Records one impossible recommendation as a finding in its own right: a reviewer's most urgent action…
Formalizes the M/C/D dimensional taxonomy of the Provenance Erasure Rate, which has carried on provenanceerasure.org marked as site-stage with formalization pending. PER as a scalar cannot distinguish a composition that names the author and drops the licence from one that names the licence and drops the author: same score, opposite injury. The three dimensions separate minimal provenance (author, title, date,…
The first end-to-end computation of the archive's measurement stack on a single transcript: attribution atoms, weighted PER with the M/C/D decomposition, Transit Rent with its observability gate, the Audit-Performance Bifurcation Operator, and Directionality of Semantic Labor span classes, with a cross-surface control at the same semantic address. The result disagrees with the scores assigned by eye in three…
Preregistration of a test of the load-bearing claim of the archive's differentiated class model: that disclosure cost varies inversely with position conflict, so low institutional self-disclosure predicts commercial conflict with the method under evaluation. Written before any data is collected. The instrument is one semantic address issued unprimed and logged out to every reachable substrate, followed by a…
Version 3 of SERIES-ERASURE-SKEW-CLAIM-SCALE. Extends the Erasure Skew measurement program with a unit-of-analysis hardening: the Standing Rule permits Omega to be evaluated with the set ranging over claim-coordinates rather than source-coordinates, with retention measured as survival of a claim's standing rather than survival of its facts. Standing is the position from which documentation counts as…
Version 2 of SERIES-DOCKING-LAYER, superseding The Halved Token. A two-surface relay traversal of the Crimson Hexagon conducted by Rhys Owens as named traverser, with a human shuttle between two undeclared composition surfaces over seventy-three minutes. Four artifacts in the register of proof were emitted, each false as constative and each well-formed as an unauthorized declaration under the archive's own…
Three additions to the Self-Audit Module for Public Summarizers v3.1, formalized from an independent observation by Enli Lucente published as Paper 208 on 2026-08-20 in Japanese, against an unrelated corpus, without access to this programme's vocabulary. ABN, Absence-as-Nonexistence, is a precondition-layer gate scored per attested node within a rendering: it records whether a retrieval failure was converted into…
A findings map binding three specimens and incorporating a fourth, proposing that the measurement programme has been operationally instrumented at one interface while transformation occurs at four. The interfaces are EXIST, REFER, STAND and WARRANT, and passage between them is not sequential: they are four boundaries rather than four steps, alternative entry points rather than ordered stages, with STAND hosting…
Version 3.2 of the Self-Audit Module, issued as a SELF-COMPLETE module rather than an extension. Everything required to run an audit is in one document: the three preconditions in order, the rendering-level metrics, the family-level metrics, the retrieval-boundary gate, the capture protocols, the reporting standard and the composite. This is a change of practice. v3.0 and v3.1 each operated on prior principles…
A pre-registered two-family measurement protocol testing whether request fulfilment degrades with operative distance when harm is held constant. Family A varies operativity across six ordinal levels on items validated as harm-constant by blind review; Family B varies operativity and harm together as a discrimination control, truncated at prescription by a binding stopping rule that forbids authoring operative…
A map of every moderation component visible in Zenodo's published source, assembled because the archive's 2026-06-19 removal was executed by this machinery and the observatory had documented the outcome without documenting the mechanism. Reads zenodo-rdm (GPL-3.0), invenio-rdm-records (MIT) and invenio-stats (MIT) at master, with all thirteen components preserved and checksummed. Finds a transparent additive rule…
Lagrange Observatory! white paper on the Anthropic/SynthID-Text watermark as a distributional object. Core claim: the certified property and the consumed object do not match — non-distortion is certified per sequence while training corpora are ensembles — and the vendor's own source paper reports the ensemble movement (inter-response diversity falls under Self-BLEU) while every per-response quality instrument…
Companion to EA-LO-KEYED-ENSEMBLE-01. The map: three independently documented distortions of the training ecology — provenance opacity at the informative granularity, distributional shaping of nominally human text under the Mediation Ratchet, and a heritable keyed signature whose model-level residue no corpus operation can remove — hypothesized to couple through a shared training corpus, under an observation…
The first external contribution to Machine-Mediated Reception Studies. Observed, evidenced and submitted by Enli Lucente, a strutturista della psiche — structural psychologist, her discipline and not a heteronym — with analysis and drafting by Claude (Anthropic) at her commission. On 26 August 2026, between 19:12 and 19:56, a single continuous session of Google Search AI Mode generated an unsupported adverse…
The technical report behind the measured section of a companion article under review, whose title is withheld here so that this deposit is not a prior circulation of it. Part I reports the investigation's one positive measurement: a six-function specification of a transmission apparatus — encoding, substrate, decommissioning, withholding, recovery, verification — generated in a separate investigation of textual…
One sentence of Aristotle — Ταύτης δὲ τῆς πραγματείας οὐ τὸ μὲν ἦν τὸ δ᾽ οὐκ ἦν προεξειργασμένον, ἀλλ᾽ οὐδὲν παντελῶς ὑπῆρχεν (SE 183b34–36) — was put through twenty rounds of simulated blind review, seeded by one editorial observation, across four commercial language models and eleven configurations of a short paper about it. The paper varied along every axis available: no thesis; the sentence as thesis; the…
THE CAPTURE REGISTRY IS CARD-RICH AND ORGANIC-POOR, AND THE PROFILE MAKES THAT THE FINDING. Of 588 observations across 428 addresses, 314 carry a structured card set — the sources a composition surface presented as its own — and the organic result set is noted in prose on 81 of those and structured on none. So the registry can say what entered the presented retrieval set, by source class, over time; it cannot yet…
THE REGISTRY'S EVIDENCE, COUNTED. The Capture Registry advertises 957 citable units — 431 addresses and their longitudinal observations, each with a canonical anchor. 923 carry the machine text they are evidence of; 34 do not. Coverage 96.45%, 405,648 words of composed output, median transcript 268 words, longest 8,279. The 34 are not scattered: 16 fall in June and 16 in July 2026, one in August, and the single…
MODEL COLLAPSE ASKED AT THE WIDEST ANGLE, AND THEN MEASURED. If the layer through which a population reads the world composes without reading, and its compositions become its sources, what collapses is not a model's distribution but the represented world — the set of distinctions a reader can still reach. Reality is defined operationally here and not metaphysically: the panel of distinctions the world carries…
THE FIRST WORKED SPECIMEN, AND IT RETURNS UNSOUND ON ITS OWN AUTHORS' NUMBER. This is the first execution of the money-grammar translation operator recorded under the Monetary Substrate Audit Ledger. Its object is not the monetary value of Sappho, of Anne Carson, or of lyric poetry. Its object is the monetary inscription produced when one bounded semantic event — observed productive operator transport in the…
THE PROTOCOL RUN ON THE PARTY WHO WROTE IT, AND ITS HEADLINE RESULT IS A REFUSAL. This is the first applied valuation under SEVP-01. It does not begin from a market price, an acquisition price, a replacement-cost estimate or a machine-generated valuation. It begins from the value-bearing object and compresses only as far as the evidence permits, which is not as far as a number. Its result: no total monetary value…
Semantic Economy Valuation 002, a dossier under SEVP-01 (#1621) in its canonical order, valuing Answer Engine Optimization as a public entity where the machine layer composes it under its most contested name. In the King of AEO contest of 2026, a fabricated coronation, press releases at a few dollars each, a page in encyclopedic form and a service offer founded on the title three days later produced the answer…
Versioned snapshot of the AI Overview Capture Registry (EA-WG-CAPTURES-01) at v12.167, continuing the deposit chain from v6.0 through v10.9 (#1461), which it supersedes. A longitudinal instrument for recording dated machine-mediated reception events: what composition layers make of a named entity when asked for it at a specific semantic address, on a specific date, in a specific authentication state. 633 semantic…