{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:OQ6NAHCKRJLJLGR5AFCMVL654E","short_pith_number":"pith:OQ6NAHCK","schema_version":"1.0","canonical_sha256":"743cd01c4a8a56959a3d0144caafdde11b4a773f8545644b99a3cc0010ce19b7","source":{"kind":"arxiv","id":"2411.02481","version":3},"attestation_state":"computed","paper":{"title":"Dr. SoW: Density Ratio of Strong-over-weak LLMs for Reducing the Cost of Human Annotation in Preference Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Akash Srivastava, Guangxuan Xu, Hao Wang, Kai Xu, Shivchander Sudalairaj","submitted_at":"2024-11-04T18:54:39Z","abstract_excerpt":"Preference tuning relies on high-quality human preference data, which is often expensive and time-consuming to gather. In this paper, we introduce Dr.SoW (Density Ratio of Strong over Weak) a cost-effective method that eliminates the reliance for human annotation by leveraging off-the-shelf LLMs for preference data annotation. Dr.SoW uses the log-density ratio between a better-aligned and a less-aligned LLM as a reward signal. We evaluate Dr.SoW across 221 different LLM pairs and empirically find a strong correlation between the performance gap of the paired models and the quality of the rewar"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2411.02481","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-04T18:54:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a19a48a8a0e891332b94a730bc46f50180ac01f31c5e42edfd20558979d8d683","abstract_canon_sha256":"b7296cc44e4101e97138faff74689e341bdb157cd35cfb8665876e5a1e9deadf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:08:24.253490Z","signature_b64":"+PuQ2ZKwOqERVENgTt0fV0GYoORHeWyTwlmlk/hKGPKBdSRSb2XIgSoz2vue36zDeHUJVf/8hFJIwd93VhKSBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"743cd01c4a8a56959a3d0144caafdde11b4a773f8545644b99a3cc0010ce19b7","last_reissued_at":"2026-07-05T10:08:24.253043Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:08:24.253043Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dr. SoW: Density Ratio of Strong-over-weak LLMs for Reducing the Cost of Human Annotation in Preference Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Akash Srivastava, Guangxuan Xu, Hao Wang, Kai Xu, Shivchander Sudalairaj","submitted_at":"2024-11-04T18:54:39Z","abstract_excerpt":"Preference tuning relies on high-quality human preference data, which is often expensive and time-consuming to gather. In this paper, we introduce Dr.SoW (Density Ratio of Strong over Weak) a cost-effective method that eliminates the reliance for human annotation by leveraging off-the-shelf LLMs for preference data annotation. Dr.SoW uses the log-density ratio between a better-aligned and a less-aligned LLM as a reward signal. We evaluate Dr.SoW across 221 different LLM pairs and empirically find a strong correlation between the performance gap of the paired models and the quality of the rewar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02481","kind":"arxiv","version":3},"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/2411.02481/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2411.02481","created_at":"2026-07-05T10:08:24.253097+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.02481v3","created_at":"2026-07-05T10:08:24.253097+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02481","created_at":"2026-07-05T10:08:24.253097+00:00"},{"alias_kind":"pith_short_12","alias_value":"OQ6NAHCKRJLJ","created_at":"2026-07-05T10:08:24.253097+00:00"},{"alias_kind":"pith_short_16","alias_value":"OQ6NAHCKRJLJLGR5","created_at":"2026-07-05T10:08:24.253097+00:00"},{"alias_kind":"pith_short_8","alias_value":"OQ6NAHCK","created_at":"2026-07-05T10:08:24.253097+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E","json":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E.json","graph_json":"https://pith.science/api/pith-number/OQ6NAHCKRJLJLGR5AFCMVL654E/graph.json","events_json":"https://pith.science/api/pith-number/OQ6NAHCKRJLJLGR5AFCMVL654E/events.json","paper":"https://pith.science/paper/OQ6NAHCK"},"agent_actions":{"view_html":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E","download_json":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E.json","view_paper":"https://pith.science/paper/OQ6NAHCK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.02481&json=true","fetch_graph":"https://pith.science/api/pith-number/OQ6NAHCKRJLJLGR5AFCMVL654E/graph.json","fetch_events":"https://pith.science/api/pith-number/OQ6NAHCKRJLJLGR5AFCMVL654E/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E/action/storage_attestation","attest_author":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E/action/author_attestation","sign_citation":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E/action/citation_signature","submit_replication":"https://pith.science/pith/OQ6NAHCKRJLJLGR5AFCMVL654E/action/replication_record"}},"created_at":"2026-07-05T10:08:24.253097+00:00","updated_at":"2026-07-05T10:08:24.253097+00:00"}