{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GFCOB7U2ZC5AXDWNJ75UFJHGPT","short_pith_number":"pith:GFCOB7U2","schema_version":"1.0","canonical_sha256":"3144e0fe9ac8ba0b8ecd4ffb42a4e67cc96a44b6fcf446b1dfb858d5ebeda869","source":{"kind":"arxiv","id":"2502.04354","version":1},"attestation_state":"computed","paper":{"title":"Reviving The Classics: Active Reward Modeling in Large Language Model Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Hao Sun, Jean-Fran\\c{c}ois Ton, Yunyi Shen","submitted_at":"2025-02-04T18:47:11Z","abstract_excerpt":"Building neural reward models from human preferences is a pivotal component in reinforcement learning from human feedback (RLHF) and large language model alignment research. Given the scarcity and high cost of human annotation, how to select the most informative pairs to annotate is an essential yet challenging open problem. In this work, we highlight the insight that an ideal comparison dataset for reward modeling should balance exploration of the representation space and make informative comparisons between pairs with moderate reward differences. Technically, challenges arise in quantifying "},"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":"2502.04354","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-04T18:47:11Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"64c3d3247060000a146c289e20a36dd2e1d5dcef11c154d90fd68fe01dfe4030","abstract_canon_sha256":"c08e5c777593b618b4b7b6b1c13762e44f3c499e596c89aaeb45c887cc9c1ef0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:50.182180Z","signature_b64":"I61RgPEV/y+A6XRo0nrjTt4qOuy0FrHy3/zPAtOTQc+JJaIZK95+Xw0iP5Mi3vC0vI1qijWvTpaOVWNM5cj1Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3144e0fe9ac8ba0b8ecd4ffb42a4e67cc96a44b6fcf446b1dfb858d5ebeda869","last_reissued_at":"2026-07-05T10:10:50.181752Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:50.181752Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reviving The Classics: Active Reward Modeling in Large Language Model Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Hao Sun, Jean-Fran\\c{c}ois Ton, Yunyi Shen","submitted_at":"2025-02-04T18:47:11Z","abstract_excerpt":"Building neural reward models from human preferences is a pivotal component in reinforcement learning from human feedback (RLHF) and large language model alignment research. Given the scarcity and high cost of human annotation, how to select the most informative pairs to annotate is an essential yet challenging open problem. In this work, we highlight the insight that an ideal comparison dataset for reward modeling should balance exploration of the representation space and make informative comparisons between pairs with moderate reward differences. Technically, challenges arise in quantifying "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04354","kind":"arxiv","version":1},"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/2502.04354/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":"2502.04354","created_at":"2026-07-05T10:10:50.181807+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.04354v1","created_at":"2026-07-05T10:10:50.181807+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04354","created_at":"2026-07-05T10:10:50.181807+00:00"},{"alias_kind":"pith_short_12","alias_value":"GFCOB7U2ZC5A","created_at":"2026-07-05T10:10:50.181807+00:00"},{"alias_kind":"pith_short_16","alias_value":"GFCOB7U2ZC5AXDWN","created_at":"2026-07-05T10:10:50.181807+00:00"},{"alias_kind":"pith_short_8","alias_value":"GFCOB7U2","created_at":"2026-07-05T10:10:50.181807+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/GFCOB7U2ZC5AXDWNJ75UFJHGPT","json":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT.json","graph_json":"https://pith.science/api/pith-number/GFCOB7U2ZC5AXDWNJ75UFJHGPT/graph.json","events_json":"https://pith.science/api/pith-number/GFCOB7U2ZC5AXDWNJ75UFJHGPT/events.json","paper":"https://pith.science/paper/GFCOB7U2"},"agent_actions":{"view_html":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT","download_json":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT.json","view_paper":"https://pith.science/paper/GFCOB7U2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.04354&json=true","fetch_graph":"https://pith.science/api/pith-number/GFCOB7U2ZC5AXDWNJ75UFJHGPT/graph.json","fetch_events":"https://pith.science/api/pith-number/GFCOB7U2ZC5AXDWNJ75UFJHGPT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT/action/storage_attestation","attest_author":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT/action/author_attestation","sign_citation":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT/action/citation_signature","submit_replication":"https://pith.science/pith/GFCOB7U2ZC5AXDWNJ75UFJHGPT/action/replication_record"}},"created_at":"2026-07-05T10:10:50.181807+00:00","updated_at":"2026-07-05T10:10:50.181807+00:00"}