{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:BAPKPM2A7PECQXIUJ2GNDZ3ILG","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":"671bd616d612963c99f15394ca187ce16429a5b3842c8accfe57b3c3aeee3565","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-03T05:46:53Z","title_canon_sha256":"d74ff69db6e93de39b571e17fa53610dedafa980d90cf5628bb600d0cf0fb5f8"},"schema_version":"1.0","source":{"id":"2406.01013","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.01013","created_at":"2026-07-05T08:34:19Z"},{"alias_kind":"arxiv_version","alias_value":"2406.01013v2","created_at":"2026-07-05T08:34:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01013","created_at":"2026-07-05T08:34:19Z"},{"alias_kind":"pith_short_12","alias_value":"BAPKPM2A7PEC","created_at":"2026-07-05T08:34:19Z"},{"alias_kind":"pith_short_16","alias_value":"BAPKPM2A7PECQXIU","created_at":"2026-07-05T08:34:19Z"},{"alias_kind":"pith_short_8","alias_value":"BAPKPM2A","created_at":"2026-07-05T08:34:19Z"}],"graph_snapshots":[{"event_id":"sha256:e7be50246cb3e924e59bb1045df1c5d85d730fdbb69959a5c6c1cc2a79694437","target":"graph","created_at":"2026-07-05T08:34:19Z","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/2406.01013/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 enabled significant advancements within language modeling for powerful, instruction-following models. However, the alignment of these models remains a pressing challenge as the policy tends to overfit the learned ``proxy\" reward model past an inflection point of utility as measured by a ``gold\" reward model that is more performant -- a phenomenon known as overoptimisation. Prior work has mitigated this issue by computing a pessimistic statistic over an ensemble of reward models, which is common in Offline Reinforcement Learning but incredib","authors_text":"Ahmed M. Ahmed, Rafael Rafailov, Sanmi Koyejo, Stepan Sharkov, Xuechen Li","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-03T05:46:53Z","title":"Scalable Ensembling For Mitigating Reward Overoptimisation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01013","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:fe37b850b363a1c381954d7da178e54dce0caa678e5518b97c0977e03286b331","target":"record","created_at":"2026-07-05T08:34:19Z","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":"671bd616d612963c99f15394ca187ce16429a5b3842c8accfe57b3c3aeee3565","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-03T05:46:53Z","title_canon_sha256":"d74ff69db6e93de39b571e17fa53610dedafa980d90cf5628bb600d0cf0fb5f8"},"schema_version":"1.0","source":{"id":"2406.01013","kind":"arxiv","version":2}},"canonical_sha256":"081ea7b340fbc8285d144e8cd1e76859a3df7156ccf9fd59c8abc886c6fc39db","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"081ea7b340fbc8285d144e8cd1e76859a3df7156ccf9fd59c8abc886c6fc39db","first_computed_at":"2026-07-05T08:34:19.965720Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:34:19.965720Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"k2vBDLnkxHij6SM1BDWHEsR1TuMEh3PhtzpXeN0WsJViDkrebYVkrXgSFcygP62QbQzKMpoEcACCANTfmKynBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:34:19.966174Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.01013","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fe37b850b363a1c381954d7da178e54dce0caa678e5518b97c0977e03286b331","sha256:e7be50246cb3e924e59bb1045df1c5d85d730fdbb69959a5c6c1cc2a79694437"],"state_sha256":"2c354f9f0348f235bbae445bf7374b6eaa2e99cf1267800ce8c6bab430f2ba13"}