{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:AKUOQVQN52P2DJ67E2FONQFX7Q","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":"76aff074b5659055ae86044c4066f2650493af99142ff8971d6ed9d78d439499","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T18:10:53Z","title_canon_sha256":"5d44f02811377d4cb6672388eeeb2f70ccf2255126fa1f5b1d4a7aede10fd668"},"schema_version":"1.0","source":{"id":"2501.19358","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.19358","created_at":"2026-07-05T11:14:03Z"},{"alias_kind":"arxiv_version","alias_value":"2501.19358v3","created_at":"2026-07-05T11:14:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.19358","created_at":"2026-07-05T11:14:03Z"},{"alias_kind":"pith_short_12","alias_value":"AKUOQVQN52P2","created_at":"2026-07-05T11:14:03Z"},{"alias_kind":"pith_short_16","alias_value":"AKUOQVQN52P2DJ67","created_at":"2026-07-05T11:14:03Z"},{"alias_kind":"pith_short_8","alias_value":"AKUOQVQN","created_at":"2026-07-05T11:14:03Z"}],"graph_snapshots":[{"event_id":"sha256:27efa2cd7141183de9b2fc96dc2a97f287ea4695b6efef9c012aa991632f178b","target":"graph","created_at":"2026-07-05T11:14:03Z","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/2501.19358/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This work identifies the Energy Loss Phenomenon in Reinforcement Learning from Human Feedback (RLHF) and its connection to reward hacking. Specifically, energy loss in the final layer of a Large Language Model (LLM) gradually increases during the RL process, with an excessive increase in energy loss characterizing reward hacking. Beyond empirical analysis, we further provide a theoretical foundation by proving that, under mild conditions, the increased energy loss reduces the upper bound of contextual relevance in LLMs, which is a critical aspect of reward hacking as the reduced contextual rel","authors_text":"Dacheng Tao, Lefei Zhang, Liang Ding, Sen Zhang, Yuchun Miao, Yuqi Zhang","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T18:10:53Z","title":"The Energy Loss Phenomenon in RLHF: A New Perspective on Mitigating Reward Hacking"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.19358","kind":"arxiv","version":3},"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":"sha256:03029d4abd6d83321b86c882e347ae922a25c6ad5f208ea3f16b2832c42d3d93","target":"record","created_at":"2026-07-05T11:14:03Z","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":"76aff074b5659055ae86044c4066f2650493af99142ff8971d6ed9d78d439499","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T18:10:53Z","title_canon_sha256":"5d44f02811377d4cb6672388eeeb2f70ccf2255126fa1f5b1d4a7aede10fd668"},"schema_version":"1.0","source":{"id":"2501.19358","kind":"arxiv","version":3}},"canonical_sha256":"02a8e8560dee9fa1a7df268ae6c0b7fc0c650d2721f7bf85dec6e2d0747d48d0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"02a8e8560dee9fa1a7df268ae6c0b7fc0c650d2721f7bf85dec6e2d0747d48d0","first_computed_at":"2026-07-05T11:14:03.018222Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:14:03.018222Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JMWcqfnbkwmbsv3n6H6BfcdY4qTU+T4J1sZdc/NM+jCs/10H1xO2oz7HPRyMYmzR25P1Y+W78tSHF20o2ZfFCw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:14:03.018776Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.19358","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:03029d4abd6d83321b86c882e347ae922a25c6ad5f208ea3f16b2832c42d3d93","sha256:27efa2cd7141183de9b2fc96dc2a97f287ea4695b6efef9c012aa991632f178b"],"state_sha256":"10be92ef94300501cddea7dd01e35f869bf091825c3bc52ad83c3300afbbc13d"}