{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:W6UHQ6VVNDCZPFOSU4WMVIZCFZ","short_pith_number":"pith:W6UHQ6VV","schema_version":"1.0","canonical_sha256":"b7a8787ab568c59795d2a72ccaa3222e75ed189c7d828b101d5bff1f5dab425e","source":{"kind":"arxiv","id":"2310.15694","version":5},"attestation_state":"computed","paper":{"title":"COPR: Continual Learning Human Preference through Optimal Policy Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Han Zhang, Hui Wang, Lin Gui, Ruifeng Xu, Yuanzhao Zhai, Yu Lei","submitted_at":"2023-10-24T10:05:32Z","abstract_excerpt":"The technique of Reinforcement Learning from Human Feedback (RLHF) is a commonly employed method to improve pre-trained Language Models (LM), enhancing their ability to conform to human preferences. Nevertheless, the current RLHF-based LMs necessitate full retraining each time novel queries or feedback are introduced, which becomes a challenging task because human preferences can vary between different domains or tasks. Retraining LMs poses practical difficulties in many real-world situations due to the significant time and computational resources required, along with concerns related to data "},"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":"2310.15694","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-24T10:05:32Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"1621fad34099b3ff6152acb1d28d43264b245014040912791b829f342b7f868b","abstract_canon_sha256":"f8209cabd163b30c059f55a96401c95459ad103653fa5f2f246c44976b3ba4de"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:00:37.173744Z","signature_b64":"HjQB+NJCF7ghGL07OfwZp+w9CcdMhAf0UJiRXh2KgWBRZul/2uDP1SWr4RGPykeMi1V5v2pbfGYNHCBc2j7qCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b7a8787ab568c59795d2a72ccaa3222e75ed189c7d828b101d5bff1f5dab425e","last_reissued_at":"2026-07-05T08:00:37.173267Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:00:37.173267Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"COPR: Continual Learning Human Preference through Optimal Policy Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Han Zhang, Hui Wang, Lin Gui, Ruifeng Xu, Yuanzhao Zhai, Yu Lei","submitted_at":"2023-10-24T10:05:32Z","abstract_excerpt":"The technique of Reinforcement Learning from Human Feedback (RLHF) is a commonly employed method to improve pre-trained Language Models (LM), enhancing their ability to conform to human preferences. Nevertheless, the current RLHF-based LMs necessitate full retraining each time novel queries or feedback are introduced, which becomes a challenging task because human preferences can vary between different domains or tasks. Retraining LMs poses practical difficulties in many real-world situations due to the significant time and computational resources required, along with concerns related to data "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.15694","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/2310.15694/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":"2310.15694","created_at":"2026-07-05T08:00:37.173325+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.15694v5","created_at":"2026-07-05T08:00:37.173325+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.15694","created_at":"2026-07-05T08:00:37.173325+00:00"},{"alias_kind":"pith_short_12","alias_value":"W6UHQ6VVNDCZ","created_at":"2026-07-05T08:00:37.173325+00:00"},{"alias_kind":"pith_short_16","alias_value":"W6UHQ6VVNDCZPFOS","created_at":"2026-07-05T08:00:37.173325+00:00"},{"alias_kind":"pith_short_8","alias_value":"W6UHQ6VV","created_at":"2026-07-05T08:00:37.173325+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.05285","citing_title":"Attribution-Guided Continual Learning for Large Language Models","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ","json":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ.json","graph_json":"https://pith.science/api/pith-number/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/graph.json","events_json":"https://pith.science/api/pith-number/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/events.json","paper":"https://pith.science/paper/W6UHQ6VV"},"agent_actions":{"view_html":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ","download_json":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ.json","view_paper":"https://pith.science/paper/W6UHQ6VV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.15694&json=true","fetch_graph":"https://pith.science/api/pith-number/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/graph.json","fetch_events":"https://pith.science/api/pith-number/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/action/storage_attestation","attest_author":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/action/author_attestation","sign_citation":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/action/citation_signature","submit_replication":"https://pith.science/pith/W6UHQ6VVNDCZPFOSU4WMVIZCFZ/action/replication_record"}},"created_at":"2026-07-05T08:00:37.173325+00:00","updated_at":"2026-07-05T08:00:37.173325+00:00"}