{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:EMX4ERL7KEIMZVWXHSMI2MQL3X","short_pith_number":"pith:EMX4ERL7","canonical_record":{"source":{"id":"2403.19159","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T06:03:47Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ebc0df157ee15775030357d9ced5924cfc14d10983aa0c0d39eff49d4f590850","abstract_canon_sha256":"1b0e6fb56f2466560089af4054cf9eb1d31d6ec4d03ee7fb5a5d7177817b128b"},"schema_version":"1.0"},"canonical_sha256":"232fc2457f5110ccd6d73c988d320bddd66799dbbfd5d648880c39758130bcdb","source":{"kind":"arxiv","id":"2403.19159","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.19159","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"arxiv_version","alias_value":"2403.19159v2","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.19159","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"pith_short_12","alias_value":"EMX4ERL7KEIM","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"pith_short_16","alias_value":"EMX4ERL7KEIMZVWX","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"pith_short_8","alias_value":"EMX4ERL7","created_at":"2026-07-05T09:04:34Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:EMX4ERL7KEIMZVWXHSMI2MQL3X","target":"record","payload":{"canonical_record":{"source":{"id":"2403.19159","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T06:03:47Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ebc0df157ee15775030357d9ced5924cfc14d10983aa0c0d39eff49d4f590850","abstract_canon_sha256":"1b0e6fb56f2466560089af4054cf9eb1d31d6ec4d03ee7fb5a5d7177817b128b"},"schema_version":"1.0"},"canonical_sha256":"232fc2457f5110ccd6d73c988d320bddd66799dbbfd5d648880c39758130bcdb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:04:34.094672Z","signature_b64":"XkJ0BcxZTrXGJJWMWnld+LxPjRKpD6JeDPFXyKu+7PDnPxLAe2W8vs0GFsl2lhBtbPgEEQ4YNoS8aPJhc3PbDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"232fc2457f5110ccd6d73c988d320bddd66799dbbfd5d648880c39758130bcdb","last_reissued_at":"2026-07-05T09:04:34.094134Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:04:34.094134Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.19159","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":"2026-07-05T09:04:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KOvWhZcS8Y3Sgb7wLoWLrfL6pxxKGlERPNtBe14Bw2EApU6fP0TLl1dZ3kYyz88Jo4Q1eCBvEPLE++O70OKWCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T17:39:46.000400Z"},"content_sha256":"9f9f26aa03afdbc5022b9109c5436cbe178b7960848513deb10502050d8caccf","schema_version":"1.0","event_id":"sha256:9f9f26aa03afdbc5022b9109c5436cbe178b7960848513deb10502050d8caccf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:EMX4ERL7KEIMZVWXHSMI2MQL3X","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Disentangling Length from Quality in Direct Preference Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chelsea Finn, Rafael Rafailov, Ryan Park, Stefano Ermon","submitted_at":"2024-03-28T06:03:47Z","abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) has been a crucial component in the recent success of Large Language Models. However, RLHF is know to exploit biases in human preferences, such as verbosity. A well-formatted and eloquent answer is often more highly rated by users, even when it is less helpful and objective. A number of approaches have been developed to control those biases in the classical RLHF literature, but the problem remains relatively under-explored for Direct Alignment Algorithms such as Direct Preference Optimization (DPO). Unlike classical RLHF, DPO does not train a s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.19159","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/2403.19159/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":"2026-07-05T09:04:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sWPpxFoammZO2KR3bv/yd6hqgYtUdIIPpfpkGIB7WFNsoPmKUH2z2X1T3coCi4RVRP/5e9wyzlk4ZgvPiV+BCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T17:39:46.001327Z"},"content_sha256":"6bd0af137da8db6fefbd53f036892f669165387a7c1a4f01fdd32a58144b751d","schema_version":"1.0","event_id":"sha256:6bd0af137da8db6fefbd53f036892f669165387a7c1a4f01fdd32a58144b751d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X/bundle.json","state_url":"https://pith.science/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X/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":"2026-08-07T17:39:46Z","links":{"resolver":"https://pith.science/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X","bundle":"https://pith.science/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X/bundle.json","state":"https://pith.science/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EMX4ERL7KEIMZVWXHSMI2MQL3X/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:EMX4ERL7KEIMZVWXHSMI2MQL3X","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":"1b0e6fb56f2466560089af4054cf9eb1d31d6ec4d03ee7fb5a5d7177817b128b","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T06:03:47Z","title_canon_sha256":"ebc0df157ee15775030357d9ced5924cfc14d10983aa0c0d39eff49d4f590850"},"schema_version":"1.0","source":{"id":"2403.19159","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.19159","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"arxiv_version","alias_value":"2403.19159v2","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.19159","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"pith_short_12","alias_value":"EMX4ERL7KEIM","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"pith_short_16","alias_value":"EMX4ERL7KEIMZVWX","created_at":"2026-07-05T09:04:34Z"},{"alias_kind":"pith_short_8","alias_value":"EMX4ERL7","created_at":"2026-07-05T09:04:34Z"}],"graph_snapshots":[{"event_id":"sha256:6bd0af137da8db6fefbd53f036892f669165387a7c1a4f01fdd32a58144b751d","target":"graph","created_at":"2026-07-05T09:04:34Z","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/2403.19159/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) has been a crucial component in the recent success of Large Language Models. However, RLHF is know to exploit biases in human preferences, such as verbosity. A well-formatted and eloquent answer is often more highly rated by users, even when it is less helpful and objective. A number of approaches have been developed to control those biases in the classical RLHF literature, but the problem remains relatively under-explored for Direct Alignment Algorithms such as Direct Preference Optimization (DPO). Unlike classical RLHF, DPO does not train a s","authors_text":"Chelsea Finn, Rafael Rafailov, Ryan Park, Stefano Ermon","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T06:03:47Z","title":"Disentangling Length from Quality in Direct Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.19159","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":"sha256:9f9f26aa03afdbc5022b9109c5436cbe178b7960848513deb10502050d8caccf","target":"record","created_at":"2026-07-05T09:04:34Z","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":"1b0e6fb56f2466560089af4054cf9eb1d31d6ec4d03ee7fb5a5d7177817b128b","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T06:03:47Z","title_canon_sha256":"ebc0df157ee15775030357d9ced5924cfc14d10983aa0c0d39eff49d4f590850"},"schema_version":"1.0","source":{"id":"2403.19159","kind":"arxiv","version":2}},"canonical_sha256":"232fc2457f5110ccd6d73c988d320bddd66799dbbfd5d648880c39758130bcdb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"232fc2457f5110ccd6d73c988d320bddd66799dbbfd5d648880c39758130bcdb","first_computed_at":"2026-07-05T09:04:34.094134Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:04:34.094134Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"XkJ0BcxZTrXGJJWMWnld+LxPjRKpD6JeDPFXyKu+7PDnPxLAe2W8vs0GFsl2lhBtbPgEEQ4YNoS8aPJhc3PbDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:04:34.094672Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.19159","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9f9f26aa03afdbc5022b9109c5436cbe178b7960848513deb10502050d8caccf","sha256:6bd0af137da8db6fefbd53f036892f669165387a7c1a4f01fdd32a58144b751d"],"state_sha256":"8cc01eed22e6d9c392be256096aa6c1f9634b5572d9910ab1c041891cc404c8d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LHvW+v5avSCODapmWtjIiheYkyBniKif7Fqqx3BgDSDkwvldAYLoIIN0uQavTA2hV62PFYMrg6m4DwBDHbqvCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T17:39:46.008427Z","bundle_sha256":"a95a27d2ccfe385e1121a47e499fb7506864ec1d00f9b3c8ef15d8879b7efd82"}}