{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:U2VLUP4WYX75TZCLO4AYLQAYEU","short_pith_number":"pith:U2VLUP4W","canonical_record":{"source":{"id":"2501.08156","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-14T14:31:45Z","cross_cats_sorted":[],"title_canon_sha256":"244e40bc05c14c37035f8bbb6e6019a431a70a47c6a25885067c15048cc1f379","abstract_canon_sha256":"ed9546aaaa47993e1b5560442951fd60ae7332aafff6d2980df2946202529c82"},"schema_version":"1.0"},"canonical_sha256":"a6aaba3f96c5ffd9e44b770185c018252b88f1ec52d43fbc533e960b6dcb0cfa","source":{"kind":"arxiv","id":"2501.08156","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.08156","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"arxiv_version","alias_value":"2501.08156v5","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.08156","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_12","alias_value":"U2VLUP4WYX75","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_16","alias_value":"U2VLUP4WYX75TZCL","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_8","alias_value":"U2VLUP4W","created_at":"2026-07-05T11:37:14Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:U2VLUP4WYX75TZCLO4AYLQAYEU","target":"record","payload":{"canonical_record":{"source":{"id":"2501.08156","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-14T14:31:45Z","cross_cats_sorted":[],"title_canon_sha256":"244e40bc05c14c37035f8bbb6e6019a431a70a47c6a25885067c15048cc1f379","abstract_canon_sha256":"ed9546aaaa47993e1b5560442951fd60ae7332aafff6d2980df2946202529c82"},"schema_version":"1.0"},"canonical_sha256":"a6aaba3f96c5ffd9e44b770185c018252b88f1ec52d43fbc533e960b6dcb0cfa","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:14.906739Z","signature_b64":"JDbI4rg3U76s+IWAfCgrFpOIcqoYFxU/zxp2SbpKQ5A7qmD1h7ErZyuCYX9tTwlsEoOOvFO98IogNYVppLdRDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a6aaba3f96c5ffd9e44b770185c018252b88f1ec52d43fbc533e960b6dcb0cfa","last_reissued_at":"2026-07-05T11:37:14.906124Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:14.906124Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.08156","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:37:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2mF4Azzg824X/anc96XzTMrK6P+WG/pVHSIEKl0GhyrlcECDTvc8yWRROUY7ieZ+i9PJqa/0xYNbs6SclrbJDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-17T16:40:58.615560Z"},"content_sha256":"8f890d8774469504c892254286a3ef451c4313a2c35629dc4ecbf1d82600bec3","schema_version":"1.0","event_id":"sha256:8f890d8774469504c892254286a3ef451c4313a2c35629dc4ecbf1d82600bec3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:U2VLUP4WYX75TZCLO4AYLQAYEU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Are DeepSeek R1 And Other Reasoning Models More Faithful?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"James Chua, Owain Evans","submitted_at":"2025-01-14T14:31:45Z","abstract_excerpt":"Language models trained to solve reasoning tasks via reinforcement learning have achieved striking results. We refer to these models as reasoning models. Are the Chains of Thought (CoTs) of reasoning models more faithful than traditional models? We evaluate three reasoning models (based on Qwen-2.5, Gemini-2, and DeepSeek-V3-Base) on an existing test of faithful CoT. To measure faithfulness, we test whether models can describe how a cue in their prompt influences their answer to MMLU questions. For example, when the cue \"A Stanford Professor thinks the answer is D\" is added to the prompt, mode"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.08156","kind":"arxiv","version":5},"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/2501.08156/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":"2026-07-05T11:37:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"StVmSCYtwUOSPzDadyQl8CwScxxU4m254PG08satvjeE09eHnTWWrJtB57LoTwMCQV0TDkDE9TTUwETDd7eyCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-17T16:40:58.615934Z"},"content_sha256":"b15e894cf9585feedf8e5500376f7b52de754cf2e7438482964f212a6ec0df8b","schema_version":"1.0","event_id":"sha256:b15e894cf9585feedf8e5500376f7b52de754cf2e7438482964f212a6ec0df8b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/U2VLUP4WYX75TZCLO4AYLQAYEU/bundle.json","state_url":"https://pith.science/pith/U2VLUP4WYX75TZCLO4AYLQAYEU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/U2VLUP4WYX75TZCLO4AYLQAYEU/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":"2026-07-17T16:40:58Z","links":{"resolver":"https://pith.science/pith/U2VLUP4WYX75TZCLO4AYLQAYEU","bundle":"https://pith.science/pith/U2VLUP4WYX75TZCLO4AYLQAYEU/bundle.json","state":"https://pith.science/pith/U2VLUP4WYX75TZCLO4AYLQAYEU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/U2VLUP4WYX75TZCLO4AYLQAYEU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:U2VLUP4WYX75TZCLO4AYLQAYEU","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":"ed9546aaaa47993e1b5560442951fd60ae7332aafff6d2980df2946202529c82","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-14T14:31:45Z","title_canon_sha256":"244e40bc05c14c37035f8bbb6e6019a431a70a47c6a25885067c15048cc1f379"},"schema_version":"1.0","source":{"id":"2501.08156","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.08156","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"arxiv_version","alias_value":"2501.08156v5","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.08156","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_12","alias_value":"U2VLUP4WYX75","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_16","alias_value":"U2VLUP4WYX75TZCL","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_8","alias_value":"U2VLUP4W","created_at":"2026-07-05T11:37:14Z"}],"graph_snapshots":[{"event_id":"sha256:b15e894cf9585feedf8e5500376f7b52de754cf2e7438482964f212a6ec0df8b","target":"graph","created_at":"2026-07-05T11:37:14Z","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/2501.08156/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language models trained to solve reasoning tasks via reinforcement learning have achieved striking results. We refer to these models as reasoning models. Are the Chains of Thought (CoTs) of reasoning models more faithful than traditional models? We evaluate three reasoning models (based on Qwen-2.5, Gemini-2, and DeepSeek-V3-Base) on an existing test of faithful CoT. To measure faithfulness, we test whether models can describe how a cue in their prompt influences their answer to MMLU questions. For example, when the cue \"A Stanford Professor thinks the answer is D\" is added to the prompt, mode","authors_text":"James Chua, Owain Evans","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-14T14:31:45Z","title":"Are DeepSeek R1 And Other Reasoning Models More Faithful?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.08156","kind":"arxiv","version":5},"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":"sha256:8f890d8774469504c892254286a3ef451c4313a2c35629dc4ecbf1d82600bec3","target":"record","created_at":"2026-07-05T11:37:14Z","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":"ed9546aaaa47993e1b5560442951fd60ae7332aafff6d2980df2946202529c82","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-14T14:31:45Z","title_canon_sha256":"244e40bc05c14c37035f8bbb6e6019a431a70a47c6a25885067c15048cc1f379"},"schema_version":"1.0","source":{"id":"2501.08156","kind":"arxiv","version":5}},"canonical_sha256":"a6aaba3f96c5ffd9e44b770185c018252b88f1ec52d43fbc533e960b6dcb0cfa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a6aaba3f96c5ffd9e44b770185c018252b88f1ec52d43fbc533e960b6dcb0cfa","first_computed_at":"2026-07-05T11:37:14.906124Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:37:14.906124Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JDbI4rg3U76s+IWAfCgrFpOIcqoYFxU/zxp2SbpKQ5A7qmD1h7ErZyuCYX9tTwlsEoOOvFO98IogNYVppLdRDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:37:14.906739Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.08156","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8f890d8774469504c892254286a3ef451c4313a2c35629dc4ecbf1d82600bec3","sha256:b15e894cf9585feedf8e5500376f7b52de754cf2e7438482964f212a6ec0df8b"],"state_sha256":"e1095c379db633f319f08907a0719713c9eaa51196151900516f4df866392287"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Bneh0SqrDmpZWgZ/9NvREQmbnYu5+xAiyOY/SzAIFsAzNPkRiD5QSL6mSJFsJls90BIMbHERJwCkQCM0irFVDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-17T16:40:58.618383Z","bundle_sha256":"df15da0c015264f8fa3d36b524c90e99fef4fe19d57ddff43f342d23d4760d71"}}