{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:NRF6PMWXFIFSG7LY5PKO6KP7OB","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":"0c48efa8ac8afe6d824f4e14044f45c031b347e2aa258c77633d02e1ac189a12","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-04-21T03:39:33Z","title_canon_sha256":"78e0b07e42a12b8fdb6bbdf9d5cdc7e2ada660f0ed6bf5209002e678324d3b86"},"schema_version":"1.0","source":{"id":"2504.14838","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.14838","created_at":"2026-07-05T10:51:51Z"},{"alias_kind":"arxiv_version","alias_value":"2504.14838v1","created_at":"2026-07-05T10:51:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.14838","created_at":"2026-07-05T10:51:51Z"},{"alias_kind":"pith_short_12","alias_value":"NRF6PMWXFIFS","created_at":"2026-07-05T10:51:51Z"},{"alias_kind":"pith_short_16","alias_value":"NRF6PMWXFIFSG7LY","created_at":"2026-07-05T10:51:51Z"},{"alias_kind":"pith_short_8","alias_value":"NRF6PMWX","created_at":"2026-07-05T10:51:51Z"}],"graph_snapshots":[{"event_id":"sha256:6fe130101341e350f1f67b94e7ae28fdfd65deb6c7cbf75dcaec4903a68958a5","target":"graph","created_at":"2026-07-05T10:51:51Z","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/2504.14838/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The reward model (RM) that represents human preferences plays a crucial role in optimizing the outputs of large language models (LLMs), e.g., through reinforcement learning from human feedback (RLHF) or rejection sampling. However, a long challenge for RM is its uncertain reliability, i.e., LLM outputs with higher rewards may not align with actual human preferences. Currently, there is a lack of a convincing metric to quantify the reliability of RMs. To bridge this gap, we propose the \\textit{\\underline{R}eliable at \\underline{$\\eta$}} (RETA) metric, which directly measures the reliability of ","authors_text":"Anxiang Zeng, Guangda Huzhang, Han Yu, Qingtao Yu, Xuesi Wang, Yawen Liu, Yizhou Chen, Zhiming Zhou","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-04-21T03:39:33Z","title":"Establishing Reliability Metrics for Reward Models in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.14838","kind":"arxiv","version":1},"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:f371f8255599157e1a4420a4b2cc955290b28a36b79963d434ddc272ada4734c","target":"record","created_at":"2026-07-05T10:51:51Z","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":"0c48efa8ac8afe6d824f4e14044f45c031b347e2aa258c77633d02e1ac189a12","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-04-21T03:39:33Z","title_canon_sha256":"78e0b07e42a12b8fdb6bbdf9d5cdc7e2ada660f0ed6bf5209002e678324d3b86"},"schema_version":"1.0","source":{"id":"2504.14838","kind":"arxiv","version":1}},"canonical_sha256":"6c4be7b2d72a0b237d78ebd4ef29ff705aa65f37b60c0eb69d89a98d37ffa9c1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6c4be7b2d72a0b237d78ebd4ef29ff705aa65f37b60c0eb69d89a98d37ffa9c1","first_computed_at":"2026-07-05T10:51:51.323189Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:51:51.323189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"N49YpFTWfvbN01fzUFEnqHSbhR8ny9VXo7MLfa7ZF8knWt+ziMNv30Bfjj42LGDV8yo72AlyuTs6hfW6at9FBw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:51:51.323730Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.14838","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f371f8255599157e1a4420a4b2cc955290b28a36b79963d434ddc272ada4734c","sha256:6fe130101341e350f1f67b94e7ae28fdfd65deb6c7cbf75dcaec4903a68958a5"],"state_sha256":"9d44d4307b052886f387bc5e9f3b6451c04d92c682c7b87eed07f1d13c680180"}