{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:KA3AJ3R7SHN5U446H4DU3UGYLV","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":"9769a0e9f5d34b1bb571a731e37c11f3f4439a883ed533503fa6600718246031","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-12-06T08:15:37Z","title_canon_sha256":"8d816d3a5e18d3ab18044d9a25cd4f61b492c741e110c56709e42b6e9c16f605"},"schema_version":"1.0","source":{"id":"2512.06343","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.06343","created_at":"2026-06-10T00:08:25Z"},{"alias_kind":"arxiv_version","alias_value":"2512.06343v3","created_at":"2026-06-10T00:08:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.06343","created_at":"2026-06-10T00:08:25Z"},{"alias_kind":"pith_short_12","alias_value":"KA3AJ3R7SHN5","created_at":"2026-06-10T00:08:25Z"},{"alias_kind":"pith_short_16","alias_value":"KA3AJ3R7SHN5U446","created_at":"2026-06-10T00:08:25Z"},{"alias_kind":"pith_short_8","alias_value":"KA3AJ3R7","created_at":"2026-06-10T00:08:25Z"}],"graph_snapshots":[{"event_id":"sha256:d782c9e86bd6f7a330db07b78f1aaa305900594bb84d9bda8e89cd76b864bf88","target":"graph","created_at":"2026-06-10T00:08:25Z","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/2512.06343/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reward models are central to Large Language Model (LLM) alignment within the framework of RLHF. The standard objective used in reward modeling is the Bradley-Terry (BT) loss, which learns from pairwise data consisting of chosen and rejected responses. In this work, we analyze the per-sample gradient of BT-loss and show spurious learning signals due to representation distance. In particular, BT gradient norm scales with two distinct components: (1) prediction error, reflected by the difference in predicted rewards between chosen and rejected responses, and critically, (2) representation distanc","authors_text":"Andrew Bai, Cho-Jui Hsieh, Haoyu Li, Tong Xie, Yuanhao Ban, Yunqi Hong","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-12-06T08:15:37Z","title":"When Distance Distracts: Representation Distance Bias in BT-Loss for Reward Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.06343","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:e76fc8ac59878811752d9543380b43a7bdd10ddd08253e5af002c6c4f53e4b7b","target":"record","created_at":"2026-06-10T00:08:25Z","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":"9769a0e9f5d34b1bb571a731e37c11f3f4439a883ed533503fa6600718246031","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-12-06T08:15:37Z","title_canon_sha256":"8d816d3a5e18d3ab18044d9a25cd4f61b492c741e110c56709e42b6e9c16f605"},"schema_version":"1.0","source":{"id":"2512.06343","kind":"arxiv","version":3}},"canonical_sha256":"503604ee3f91dbda739e3f074dd0d85d6928890ada2cd013a72b97bc3bdbed91","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"503604ee3f91dbda739e3f074dd0d85d6928890ada2cd013a72b97bc3bdbed91","first_computed_at":"2026-06-10T00:08:25.984224Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-10T00:08:25.984224Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"dOuaWyrV+6r+o2MoAS8fVuHSo5h7IX7QS+Gs15vb+qQvFXHswnQmH4SjhkmZECEWqFubG8R2kdopANwsJgpJAg==","signature_status":"signed_v1","signed_at":"2026-06-10T00:08:25.985209Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.06343","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e76fc8ac59878811752d9543380b43a7bdd10ddd08253e5af002c6c4f53e4b7b","sha256:d782c9e86bd6f7a330db07b78f1aaa305900594bb84d9bda8e89cd76b864bf88"],"state_sha256":"64a32f710da02e04d0434fcfdfa5d0c762a3509b016f776e2cadd53151ca32c9"}