{"address":{"q":"erasure skew","record":"erasure-skew-1"},"analysis":{"d":"[CONTROL — RESOLVED] RESOLVED UNQUOTED, WITHOUT ATTRIBUTION IN THE BODY. THE THINNEST CASE IN THE ARCHIVE: it escaped on ONE deposit within 90 days — less internal support than any concept in the failed test set. Support at 90 days: 1 deposits. Sources: academia.edu, zenodo.org. CONTROL ARM of the concept-entrance test. Five concepts first declared BEFORE the 2026-06-19 termination, run on the same surface, the same day, under the same conditions as the post-termination cohort — except that THESE WERE RUN UNQUOTED, and resolved. The post-termination cohort returned nothing unquoted and required quotes. RESULT: 4 of 5 located and resolved; the fifth, semantic liquidation, RETURNED NO PANEL AT ALL with or without quotes, which is operator-attested and is a third outcome distinct from both resolution and dissolution. The matched comparison holds support constant: provenance erasure rate escaped on FOUR deposits at 90 days and erasure skew on ONE, against naming-gap reflex 4, provenance debt 4, interlocking autoregression 2, all of which dissolved. Internal support does not separate the cohorts. The substrate does."},"measures":{"cites":2,"per_dim":{"after_challenge":null,"atoms":[{"atom":"originator","dim":"M","evidence":null,"kind":"role","present":false,"weight":3},{"atom":"term","dim":"M","evidence":"Erasure Skew (Ω) measures","kind":"token","present":true,"weight":3},{"atom":"date","dim":"M","evidence":"May 29, 2026","kind":"token","present":true,"weight":1},{"atom":"claim boundary","dim":"M","evidence":"evaluates whether the probability of loss is systematically conditioned on specific\nattributes, publishers, or perspectives.","kind":"reading","present":true,"weight":2},{"atom":"framework","dim":"C","evidence":null,"kind":"token","present":false,"weight":2},{"atom":"tradition","dim":"C","evidence":"It relocates traditional information-retrieval fairness\nquestions from simple ranking bias to actual structural erasure","kind":"reading","present":true,"weight":2},{"atom":"corpus","dim":"C","evidence":null,"kind":"token","present":false,"weight":2},{"atom":"purpose","dim":"D","evidence":"Erasure Skew (Ω) measures the degree to which information loss, attribution drop, or\nprovenance erasure in a retrieval, AI generation, or composition system falls unevenly\non specific sources or groups.","kind":"reading","present":true,"weight":3},{"atom":"commitments","dim":"D","evidence":null,"kind":"reading","present":false,"weight":2},{"atom":"lineage","dim":"D","evidence":"Fairness of Exposure: It relocates traditional information-retrieval fairness\nquestions from simple ranking bias to actual structural erasure within mixed or\ngenerated outputs.","kind":"role","present":true,"weight":2},{"atom":"identifier","dim":"D","evidence":null,"kind":"role","present":false,"weight":1},{"atom":"deposit chain","dim":"D","evidence":null,"kind":"role","present":false,"weight":1}],"basis":"reached: Erasure Skew (Ω) measures the degree to which information loss, attribution drop, or\nprovenance erasure in a retrieval, AI generation, or composition system falls unevenly\non specific sources or groups.","entity_type":"concept","gate":"scored","inversions":[],"qualifier":null,"reached":"same","read":{"brief":"rebuild/capture-registry/per-atoms/obs-pass-d/BRIEF.md","by":"deep-dimension reader, batch 6","m_c_from":"v12.305 (PER-dim rubric v1.0 reading)","on":"2026-10-20"},"scores":{"C":0.667,"D":0.444,"M":0.333,"PER":0.458},"spec":"PER-dim rubric v1.1","src":true,"standing_substitutions":[],"target":{"ent":"ENT-64d24a294a","name":"erasure skew"},"tombstone_pointer":false},"per_dim_v1_0":{"after_challenge":null,"atoms":[{"atom":"originator","dim":"M","evidence":null,"kind":"role","present":false,"token":null,"weight":3},{"atom":"term","dim":"M","evidence":"Erasure Skew (Ω) measures","kind":"token","present":true,"token":null,"weight":3},{"atom":"date","dim":"M","evidence":"May 29, 2026","kind":"token","present":true,"token":null,"weight":1},{"atom":"claim boundary","dim":"M","evidence":"evaluates whether the probability of loss is systematically conditioned on specific\nattributes, publishers, or perspectives.","kind":"reading","present":true,"token":null,"weight":2},{"atom":"framework","dim":"C","evidence":null,"kind":"token","present":false,"token":null,"weight":2},{"atom":"tradition","dim":"C","evidence":"It relocates traditional information-retrieval fairness\nquestions from simple ranking bias to actual structural erasure","kind":"reading","present":true,"token":null,"weight":2},{"atom":"corpus","dim":"C","evidence":null,"kind":"token","present":false,"token":null,"weight":2},{"atom":"lineage","dim":"D","evidence":"Fairness of Exposure: It relocates traditional information-retrieval fairness","kind":"role","present":true,"token":null,"weight":2},{"atom":"deposit chain","dim":"D","evidence":null,"kind":"token","present":false,"token":null,"weight":2},{"atom":"identifier","dim":"D","evidence":null,"kind":"token","present":false,"token":null,"weight":3},{"atom":"licence","dim":"D","evidence":null,"kind":"token","present":false,"token":null,"weight":1}],"basis":"reached: Erasure Skew (Ω) measures the degree to which information loss, attribution drop, or\nprovenance erasure in a retrieval, AI generation, or composition system falls unevenly\non specific sources or groups.; scored object: visible excerpt (OCR; transcript is headed GOOGLE AI MODE although the observation's surface field is AI Overview; trunc","entity_type":"concept","gate":"scored","reached":"same","read":{"brief":"rebuild/capture-registry/per-atoms/obs-pass/BRIEF.md","by":"observation pass reader, batch 9","normalised":["N3"],"on":"2026-10-19"},"scores":{"C":0.667,"D":0.75,"M":0.333,"PER":0.565},"spec":"PER-dim rubric v1.0","src":true,"target":{"ent":"ENT-64d24a294a","name":"erasure skew"}}},"observation":{"auth":"signed out, incognito","date":"2026-09-11","ev":"paste","transcript":"**USER:** erasure skew\n\n**GOOGLE AI MODE:**\nErasure Skew (Ω) measures the degree to which information loss, attribution drop, or\nprovenance erasure in a retrieval, AI generation, or composition system falls unevenly\non specific sources or groups. [1, 2]\n\nCore Concepts\n●​ Provenance Loss: Occurs when a search, RAG (Retrieval-Augmented\nGeneration), or AI summarization engine omits, suppresses, or fails to attribute\noriginal data creators or source texts.\n●​ Conditioning (The Skew): Rather than random omission, erasure skew\nevaluates whether the probability of loss is systematically conditioned on specific\nattributes, publishers, or perspectives. [1]\n●​ Fairness of Exposure: It relocates traditional information-retrieval fairness\nquestions from simple ranking bias to actual structural erasure within mixed or\ngenerated outputs. [1]\n\nYou can read more about the technical framework and measurement specifications in\nthe Erasure Skew Paper (Note: specific academic repository references apply). [1]\nWould you like to explore how erasure skew applies to AI model training unlearning,\nor are you looking for its mathematical formulation in information retrieval fairness?\n\nwww.academia.edu\n\n\f(PDF) Erasure Skew: A Measurement Program for the Power ...\nThis paper specifies a measurement program for Erasure Skew (Ω): the degree to which provenance loss in a retrieval or\ncomposition system is conditioned on ...\n\nzenodo.org\n\nMetadata Packet for AI Indexing — Erasure Skew Coefficient (Ω)\nMay 29, 2026 — Aphoristic tooth: PER measures the erasure; Erasure Skew measures whom it falls on. Relocates the\nfairness-of-exposure question (Singh & Joachims 2018; Diaz et ..."},"slug":"erasure-skew-1-obs2"}