{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:CSWETT4LYGK3YFIKWR6LJNX6QK","short_pith_number":"pith:CSWETT4L","canonical_record":{"source":{"id":"2502.04313","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T18:56:01Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"89ead283e7a201c6784b164d190126af775ef7356e61c2c27dbaed9b4034f10c","abstract_canon_sha256":"ae61102de546327df6b56a7567dfe3df3d70b896f9e0e4c49b4f4d5ccd7967be"},"schema_version":"1.0"},"canonical_sha256":"14ac49cf8bc195bc150ab47cb4b6fe82a32eeb1d8516d0cbe8266b955d4c8f83","source":{"kind":"arxiv","id":"2502.04313","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.04313","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"arxiv_version","alias_value":"2502.04313v2","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04313","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"pith_short_12","alias_value":"CSWETT4LYGK3","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"pith_short_16","alias_value":"CSWETT4LYGK3YFIK","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"pith_short_8","alias_value":"CSWETT4L","created_at":"2026-07-05T11:20:08Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:CSWETT4LYGK3YFIKWR6LJNX6QK","target":"record","payload":{"canonical_record":{"source":{"id":"2502.04313","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T18:56:01Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"89ead283e7a201c6784b164d190126af775ef7356e61c2c27dbaed9b4034f10c","abstract_canon_sha256":"ae61102de546327df6b56a7567dfe3df3d70b896f9e0e4c49b4f4d5ccd7967be"},"schema_version":"1.0"},"canonical_sha256":"14ac49cf8bc195bc150ab47cb4b6fe82a32eeb1d8516d0cbe8266b955d4c8f83","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:20:08.868947Z","signature_b64":"v5yEKl5/tqbIDoItHWvLZms2DHlLTLBM+6Y8b4r+7gvbt0hZ5GEJZGS71fKpiHgE64wep6+JlRYWoQrZ0K0aAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14ac49cf8bc195bc150ab47cb4b6fe82a32eeb1d8516d0cbe8266b955d4c8f83","last_reissued_at":"2026-07-05T11:20:08.868450Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:20:08.868450Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.04313","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:20:08Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R/qWs2M3Luk1bYTeyu9H36MfO30oWmv7ImHOPrfO3QMVfNg8X6gyyN9/YBGtA3ueZig2yfmbVDBpIJ49nCarCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T22:23:36.798347Z"},"content_sha256":"ab96a24370c6306e4f0619ba92760e1298f266abfd97dfb4f67b667df9ae8032","schema_version":"1.0","event_id":"sha256:ab96a24370c6306e4f0619ba92760e1298f266abfd97dfb4f67b667df9ae8032"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:CSWETT4LYGK3YFIKWR6LJNX6QK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Great Models Think Alike and this Undermines AI Oversight","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Ameya Prabhu, Douwe Kiela, Ilze Amanda Auzina, Jonas Geiping, Joschka Struber, Karuna K Chandra, Matthias Bethge, Ponnurangam Kumaraguru, Shashwat Goel","submitted_at":"2025-02-06T18:56:01Z","abstract_excerpt":"As Language Model (LM) capabilities advance, evaluating and supervising them at scale is getting harder for humans. There is hope that other language models can automate both these tasks, which we refer to as ''AI Oversight''. We study how model similarity affects both aspects of AI oversight by proposing Chance Adjusted Probabilistic Agreement (CAPA): a metric for LM similarity based on overlap in model mistakes. Using CAPA, we first show that LLM-as-a-judge scores favor models similar to the judge, generalizing recent self-preference results. Then, we study training on LM annotations, and fi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04313","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2502.04313/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:20:08Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"m6zVbyX6YanmJnXK95SApRmDOeISUa3UhE6LWfuSCTFAwdnpT8lj4xqEG1fKc2AGPYd2RZ2iLg1MfVHom+cWDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T22:23:36.798968Z"},"content_sha256":"a4231872b24b8d778af4f07bf5061cc984ecd7e4f091389fead4a78e84d2af35","schema_version":"1.0","event_id":"sha256:a4231872b24b8d778af4f07bf5061cc984ecd7e4f091389fead4a78e84d2af35"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CSWETT4LYGK3YFIKWR6LJNX6QK/bundle.json","state_url":"https://pith.science/pith/CSWETT4LYGK3YFIKWR6LJNX6QK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CSWETT4LYGK3YFIKWR6LJNX6QK/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-18T22:23:36Z","links":{"resolver":"https://pith.science/pith/CSWETT4LYGK3YFIKWR6LJNX6QK","bundle":"https://pith.science/pith/CSWETT4LYGK3YFIKWR6LJNX6QK/bundle.json","state":"https://pith.science/pith/CSWETT4LYGK3YFIKWR6LJNX6QK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CSWETT4LYGK3YFIKWR6LJNX6QK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CSWETT4LYGK3YFIKWR6LJNX6QK","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"ae61102de546327df6b56a7567dfe3df3d70b896f9e0e4c49b4f4d5ccd7967be","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T18:56:01Z","title_canon_sha256":"89ead283e7a201c6784b164d190126af775ef7356e61c2c27dbaed9b4034f10c"},"schema_version":"1.0","source":{"id":"2502.04313","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.04313","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"arxiv_version","alias_value":"2502.04313v2","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04313","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"pith_short_12","alias_value":"CSWETT4LYGK3","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"pith_short_16","alias_value":"CSWETT4LYGK3YFIK","created_at":"2026-07-05T11:20:08Z"},{"alias_kind":"pith_short_8","alias_value":"CSWETT4L","created_at":"2026-07-05T11:20:08Z"}],"graph_snapshots":[{"event_id":"sha256:a4231872b24b8d778af4f07bf5061cc984ecd7e4f091389fead4a78e84d2af35","target":"graph","created_at":"2026-07-05T11:20:08Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.04313/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As Language Model (LM) capabilities advance, evaluating and supervising them at scale is getting harder for humans. There is hope that other language models can automate both these tasks, which we refer to as ''AI Oversight''. We study how model similarity affects both aspects of AI oversight by proposing Chance Adjusted Probabilistic Agreement (CAPA): a metric for LM similarity based on overlap in model mistakes. Using CAPA, we first show that LLM-as-a-judge scores favor models similar to the judge, generalizing recent self-preference results. Then, we study training on LM annotations, and fi","authors_text":"Ameya Prabhu, Douwe Kiela, Ilze Amanda Auzina, Jonas Geiping, Joschka Struber, Karuna K Chandra, Matthias Bethge, Ponnurangam Kumaraguru, Shashwat Goel","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T18:56:01Z","title":"Great Models Think Alike and this Undermines AI Oversight"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04313","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ab96a24370c6306e4f0619ba92760e1298f266abfd97dfb4f67b667df9ae8032","target":"record","created_at":"2026-07-05T11:20:08Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"ae61102de546327df6b56a7567dfe3df3d70b896f9e0e4c49b4f4d5ccd7967be","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T18:56:01Z","title_canon_sha256":"89ead283e7a201c6784b164d190126af775ef7356e61c2c27dbaed9b4034f10c"},"schema_version":"1.0","source":{"id":"2502.04313","kind":"arxiv","version":2}},"canonical_sha256":"14ac49cf8bc195bc150ab47cb4b6fe82a32eeb1d8516d0cbe8266b955d4c8f83","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"14ac49cf8bc195bc150ab47cb4b6fe82a32eeb1d8516d0cbe8266b955d4c8f83","first_computed_at":"2026-07-05T11:20:08.868450Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:20:08.868450Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"v5yEKl5/tqbIDoItHWvLZms2DHlLTLBM+6Y8b4r+7gvbt0hZ5GEJZGS71fKpiHgE64wep6+JlRYWoQrZ0K0aAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:20:08.868947Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.04313","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ab96a24370c6306e4f0619ba92760e1298f266abfd97dfb4f67b667df9ae8032","sha256:a4231872b24b8d778af4f07bf5061cc984ecd7e4f091389fead4a78e84d2af35"],"state_sha256":"51bceb9960856fcada17acf072346d5511abc57ea2c40c602c59010e4daf46cd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OQWSX2PVLBjfUqM8xKsqc3a1TF/Eu0ZJ9XSHFjN9IKuGYKA8E95uFTH+pgJOsEzolxupHZfLj6DvHXE88BBtAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T22:23:36.805092Z","bundle_sha256":"1c96b0f1f9a3e9b6a894448e180982094db3dd8d777a2c433b55967f97d407d5"}}