{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JDYNQUMRB6GUIHVTV3QFYENIBK","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":"757127d21a7178737541483338458bd784f94a1c1f9b41e39d9fd9c128812e19","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-21T21:49:02Z","title_canon_sha256":"15c0e81bc2106c4a5c11d9abeac7eb15f512b45ece5f9ccd23843970b2a97fab"},"schema_version":"1.0","source":{"id":"2506.17828","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.17828","created_at":"2026-07-05T11:31:07Z"},{"alias_kind":"arxiv_version","alias_value":"2506.17828v2","created_at":"2026-07-05T11:31:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.17828","created_at":"2026-07-05T11:31:07Z"},{"alias_kind":"pith_short_12","alias_value":"JDYNQUMRB6GU","created_at":"2026-07-05T11:31:07Z"},{"alias_kind":"pith_short_16","alias_value":"JDYNQUMRB6GUIHVT","created_at":"2026-07-05T11:31:07Z"},{"alias_kind":"pith_short_8","alias_value":"JDYNQUMR","created_at":"2026-07-05T11:31:07Z"}],"graph_snapshots":[{"event_id":"sha256:66e38eb6de3344940a0e788f77ccb22c52ec094a88bff4ca384e365fc2dcd1d6","target":"graph","created_at":"2026-07-05T11:31:07Z","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/2506.17828/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Aligning large language models (LLMs) with human preferences usually requires fine-tuning methods such as RLHF and DPO. These methods directly optimize the model parameters, so they cannot be used in test-time to improve model performance, nor are they applicable when the model weights are not accessible. In contrast, test-time methods sidestep weight updates by leveraging reward functions to guide and improve output quality. However, they incur high inference costs, and their one-shot guidance is often based on imperfect reward or value functions, leading to suboptimal outputs. In this work, ","authors_text":"Alfredo Garcia, Chenliang Li, Jiaxiang Li, Kaixiang Lin, Mingyi Hong, Siliang Zeng, Songtao Lu, Xinnan Zhang, Zhongruo Wang","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-21T21:49:02Z","title":"Aligning Frozen LLMs by Reinforcement Learning: An Iterative Reweight-then-Optimize Approach"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.17828","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":"sha256:313ee9af91b95ba5f153f7c069a76154876aaae517cf3a7d28645ab1d4fb942f","target":"record","created_at":"2026-07-05T11:31:07Z","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":"757127d21a7178737541483338458bd784f94a1c1f9b41e39d9fd9c128812e19","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-21T21:49:02Z","title_canon_sha256":"15c0e81bc2106c4a5c11d9abeac7eb15f512b45ece5f9ccd23843970b2a97fab"},"schema_version":"1.0","source":{"id":"2506.17828","kind":"arxiv","version":2}},"canonical_sha256":"48f0d851910f8d441eb3aee05c11a80a88a99fc9c39e7189a9a314fa84ffff82","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"48f0d851910f8d441eb3aee05c11a80a88a99fc9c39e7189a9a314fa84ffff82","first_computed_at":"2026-07-05T11:31:07.698840Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:31:07.698840Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hvi33OErXzGLO9bv60aTGXrywiUitS65wQ5Dj0L2nOHsBV/npxRZs3hJDCMHSrJ0/89YXLr5AI+Rq+lWKJJ0CQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:31:07.699311Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.17828","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:313ee9af91b95ba5f153f7c069a76154876aaae517cf3a7d28645ab1d4fb942f","sha256:66e38eb6de3344940a0e788f77ccb22c52ec094a88bff4ca384e365fc2dcd1d6"],"state_sha256":"8dbadaefae206fb7e99c6fbc44e95a4283f09ee75ddae74b59f8125ee53e7d63"}