{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3OXJAEW3BO75YJZ6FUPBZMUYCA","short_pith_number":"pith:3OXJAEW3","canonical_record":{"source":{"id":"2503.10965","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-14T00:21:15Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"3a096fd8507dac1e114cfbac3e33524699e3b761503dae2162ea771bdf595c35","abstract_canon_sha256":"178ff587c594d7f86229d1bd06c205ae24ffc12fc3cbc39f315406f8871509df"},"schema_version":"1.0"},"canonical_sha256":"dbae9012db0bbfdc273e2d1e1cb29810280578f2b50328202b53ffd9924e9473","source":{"kind":"arxiv","id":"2503.10965","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.10965","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"arxiv_version","alias_value":"2503.10965v2","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.10965","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"pith_short_12","alias_value":"3OXJAEW3BO75","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"pith_short_16","alias_value":"3OXJAEW3BO75YJZ6","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"pith_short_8","alias_value":"3OXJAEW3","created_at":"2026-07-05T10:40:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3OXJAEW3BO75YJZ6FUPBZMUYCA","target":"record","payload":{"canonical_record":{"source":{"id":"2503.10965","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-14T00:21:15Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"3a096fd8507dac1e114cfbac3e33524699e3b761503dae2162ea771bdf595c35","abstract_canon_sha256":"178ff587c594d7f86229d1bd06c205ae24ffc12fc3cbc39f315406f8871509df"},"schema_version":"1.0"},"canonical_sha256":"dbae9012db0bbfdc273e2d1e1cb29810280578f2b50328202b53ffd9924e9473","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:40:39.413246Z","signature_b64":"7qAUPFylyLsqgt7pob42kyBmRD+zJhcqQrmqL+l6DmQLJZK0EMDBXt0/o6Zki0MHPpyRL3hZ8RXZcupW1iSRBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbae9012db0bbfdc273e2d1e1cb29810280578f2b50328202b53ffd9924e9473","last_reissued_at":"2026-07-05T10:40:39.412749Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:40:39.412749Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.10965","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":"2026-07-05T10:40:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BJ9uq+5aXNOp+g1qscTYpRJwRESYh7RBzONg7/GMdhPj4SrpkJayLuBNgNn2TYkbZ4Dc2HkfF0zPo0LI31FUCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-31T14:16:25.016320Z"},"content_sha256":"5eb01c649773d2a68cf51e047c3d3419b6d628bca321f918275abe11ba3bd7e7","schema_version":"1.0","event_id":"sha256:5eb01c649773d2a68cf51e047c3d3419b6d628bca321f918275abe11ba3bd7e7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3OXJAEW3BO75YJZ6FUPBZMUYCA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Auditing language models for hidden objectives","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Adam Jermyn, Adam Pearce, Akbir Khan, Andy Shih, Austin Meek, Brian Chen, Carson Denison, Christopher Olah, Daniel Ziegler, Drake Thomas, Emmanuel Ameisen, Euan Ong, Evan Hubinger, Fabien Roger, Florian Dietz, Hoagy Cunningham, Jack Lindsey, Jan Kirchner, Jan Leike, Jeanne Salle, Johannes Treutlein, Jonathan Marcus, Joshua Batson, Kei Nishimura-Gasparian, Kelley Rivoire, Meg Tong, Monte MacDiarmid, Samuel Marks, Samuel R. Bowman, Satvik Golechha, Shan Carter, Siddharth Mishra-Sharma, Tim Belonax, Tom Henighan, Trenton Bricken","submitted_at":"2025-03-14T00:21:15Z","abstract_excerpt":"We study the feasibility of conducting alignment audits: investigations into whether models have undesired objectives. As a testbed, we train a language model with a hidden objective. Our training pipeline first teaches the model about exploitable errors in RLHF reward models (RMs), then trains the model to exploit some of these errors. We verify via out-of-distribution evaluations that the model generalizes to exhibit whatever behaviors it believes RMs rate highly, including ones not reinforced during training. We leverage this model to study alignment audits in two ways. First, we conduct a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.10965","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/2503.10965/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":"2026-07-05T10:40:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"apA+/cJNsaIzZO1LACfuhFWAMRyk7c6keCk3UIng/Ctf6BgFDjutlGJZ/QhTc8mYYYn+yPaI12LzUXXD91b4AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-31T14:16:25.016674Z"},"content_sha256":"125cf303e541cb7ad41a0dd5e93100044a3d8c8184cf0f4d8d95525eff3380f4","schema_version":"1.0","event_id":"sha256:125cf303e541cb7ad41a0dd5e93100044a3d8c8184cf0f4d8d95525eff3380f4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA/bundle.json","state_url":"https://pith.science/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA/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":"2026-07-31T14:16:25Z","links":{"resolver":"https://pith.science/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA","bundle":"https://pith.science/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA/bundle.json","state":"https://pith.science/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3OXJAEW3BO75YJZ6FUPBZMUYCA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3OXJAEW3BO75YJZ6FUPBZMUYCA","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":"178ff587c594d7f86229d1bd06c205ae24ffc12fc3cbc39f315406f8871509df","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-14T00:21:15Z","title_canon_sha256":"3a096fd8507dac1e114cfbac3e33524699e3b761503dae2162ea771bdf595c35"},"schema_version":"1.0","source":{"id":"2503.10965","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.10965","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"arxiv_version","alias_value":"2503.10965v2","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.10965","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"pith_short_12","alias_value":"3OXJAEW3BO75","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"pith_short_16","alias_value":"3OXJAEW3BO75YJZ6","created_at":"2026-07-05T10:40:39Z"},{"alias_kind":"pith_short_8","alias_value":"3OXJAEW3","created_at":"2026-07-05T10:40:39Z"}],"graph_snapshots":[{"event_id":"sha256:125cf303e541cb7ad41a0dd5e93100044a3d8c8184cf0f4d8d95525eff3380f4","target":"graph","created_at":"2026-07-05T10:40:39Z","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/2503.10965/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We study the feasibility of conducting alignment audits: investigations into whether models have undesired objectives. As a testbed, we train a language model with a hidden objective. Our training pipeline first teaches the model about exploitable errors in RLHF reward models (RMs), then trains the model to exploit some of these errors. We verify via out-of-distribution evaluations that the model generalizes to exhibit whatever behaviors it believes RMs rate highly, including ones not reinforced during training. We leverage this model to study alignment audits in two ways. First, we conduct a ","authors_text":"Adam Jermyn, Adam Pearce, Akbir Khan, Andy Shih, Austin Meek, Brian Chen, Carson Denison, Christopher Olah, Daniel Ziegler, Drake Thomas, Emmanuel Ameisen, Euan Ong, Evan Hubinger, Fabien Roger, Florian Dietz, Hoagy Cunningham, Jack Lindsey, Jan Kirchner, Jan Leike, Jeanne Salle, Johannes Treutlein, Jonathan Marcus, Joshua Batson, Kei Nishimura-Gasparian, Kelley Rivoire, Meg Tong, Monte MacDiarmid, Samuel Marks, Samuel R. Bowman, Satvik Golechha, Shan Carter, Siddharth Mishra-Sharma, Tim Belonax, Tom Henighan, Trenton Bricken","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-14T00:21:15Z","title":"Auditing language models for hidden objectives"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.10965","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":"sha256:5eb01c649773d2a68cf51e047c3d3419b6d628bca321f918275abe11ba3bd7e7","target":"record","created_at":"2026-07-05T10:40:39Z","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":"178ff587c594d7f86229d1bd06c205ae24ffc12fc3cbc39f315406f8871509df","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-14T00:21:15Z","title_canon_sha256":"3a096fd8507dac1e114cfbac3e33524699e3b761503dae2162ea771bdf595c35"},"schema_version":"1.0","source":{"id":"2503.10965","kind":"arxiv","version":2}},"canonical_sha256":"dbae9012db0bbfdc273e2d1e1cb29810280578f2b50328202b53ffd9924e9473","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dbae9012db0bbfdc273e2d1e1cb29810280578f2b50328202b53ffd9924e9473","first_computed_at":"2026-07-05T10:40:39.412749Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:40:39.412749Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"7qAUPFylyLsqgt7pob42kyBmRD+zJhcqQrmqL+l6DmQLJZK0EMDBXt0/o6Zki0MHPpyRL3hZ8RXZcupW1iSRBA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:40:39.413246Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.10965","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5eb01c649773d2a68cf51e047c3d3419b6d628bca321f918275abe11ba3bd7e7","sha256:125cf303e541cb7ad41a0dd5e93100044a3d8c8184cf0f4d8d95525eff3380f4"],"state_sha256":"32dbaa201ac3065079efe4f0b81d93d85f80d350ba24ea53183f45a5e83f08e7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"udKuTdxcnp3iadt8lEvPseB+88w2S/S5GSxi4kR6+lou/9vukWGr2pYNBFTb4uYMBQXmw/YN55zY0sM8H+4PCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-31T14:16:25.018965Z","bundle_sha256":"8f2306eae644746c1193436fc2d4bac85884ecdaab21efdd74a5cb2428af926a"}}