{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:H6X43XHDWAIJSECPR5PJ4LJMFJ","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":"46631051abb021ca270527031b82525a0b959a9550ed0472937f866b725e7fc7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-06T16:59:17Z","title_canon_sha256":"a2efabab29e20192724b2d3668dab8b7c859d1be77b79f786bef31770ef250fe"},"schema_version":"1.0","source":{"id":"2310.04373","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.04373","created_at":"2026-07-05T06:59:00Z"},{"alias_kind":"arxiv_version","alias_value":"2310.04373v2","created_at":"2026-07-05T06:59:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.04373","created_at":"2026-07-05T06:59:00Z"},{"alias_kind":"pith_short_12","alias_value":"H6X43XHDWAIJ","created_at":"2026-07-05T06:59:00Z"},{"alias_kind":"pith_short_16","alias_value":"H6X43XHDWAIJSECP","created_at":"2026-07-05T06:59:00Z"},{"alias_kind":"pith_short_8","alias_value":"H6X43XHD","created_at":"2026-07-05T06:59:00Z"}],"graph_snapshots":[{"event_id":"sha256:f3dd37d2ff57b2fd58df6dbc0dfa0739d1f1843d5f5ed55198e45c863c9da055","target":"graph","created_at":"2026-07-05T06:59:00Z","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/2310.04373/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models are typically aligned with human preferences by optimizing $\\textit{reward models}$ (RMs) fitted to human feedback. However, human preferences are multi-faceted, and it is increasingly common to derive reward from a composition of simpler reward models which each capture a different aspect of language quality. This itself presents a challenge, as it is difficult to appropriately weight these component RMs when combining them. Compounding this difficulty, because any RM is only a proxy for human evaluation, this process is vulnerable to $\\textit{overoptimization}$, wherein","authors_text":"Aaditya K. Singh, Anca D. Dragan, DJ Strouse, Ruslan Salakhutdinov, Stephen McAleer, Ted Moskovitz, Tuomas Sandholm","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-06T16:59:17Z","title":"Confronting Reward Model Overoptimization with Constrained RLHF"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.04373","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":"sha256:57d840563d6016ea4885f35ac701ef51c8dd9503ba5bb44fa67a70e50d02df24","target":"record","created_at":"2026-07-05T06:59:00Z","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":"46631051abb021ca270527031b82525a0b959a9550ed0472937f866b725e7fc7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-06T16:59:17Z","title_canon_sha256":"a2efabab29e20192724b2d3668dab8b7c859d1be77b79f786bef31770ef250fe"},"schema_version":"1.0","source":{"id":"2310.04373","kind":"arxiv","version":2}},"canonical_sha256":"3fafcddce3b01099104f8f5e9e2d2c2a5c6bd7fdd6202b3fda5e2990f4642373","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3fafcddce3b01099104f8f5e9e2d2c2a5c6bd7fdd6202b3fda5e2990f4642373","first_computed_at":"2026-07-05T06:59:00.578183Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:59:00.578183Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+dRKoUlJtn7MZmue2ZbkIXpQyM2l+twMQ+IPoPnLqxcvknbbl5/YFugqU23591C/VzsKOL7gv3J0cNDJd6DLDw==","signature_status":"signed_v1","signed_at":"2026-07-05T06:59:00.578663Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.04373","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:57d840563d6016ea4885f35ac701ef51c8dd9503ba5bb44fa67a70e50d02df24","sha256:f3dd37d2ff57b2fd58df6dbc0dfa0739d1f1843d5f5ed55198e45c863c9da055"],"state_sha256":"4b172bdfca8b699905d6187f9f7369c89175f077aabb71b13fcc424da6113ff5"}