{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:LLYFYPQ7KKAXCZ5CC5ET4AMJQ5","short_pith_number":"pith:LLYFYPQ7","canonical_record":{"source":{"id":"2602.07441","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-07T08:44:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"11ceda1decf65b89cb4513c31144ecd8ad1132fdbf83a6e4a8002b631649229e","abstract_canon_sha256":"dd379b69b869ef4149d7ec98a96f6edfb06f271348c99998106e1a120b549e75"},"schema_version":"1.0"},"canonical_sha256":"5af05c3e1f52817167a217493e018987643dfe7037bdd563807d036431926018","source":{"kind":"arxiv","id":"2602.07441","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.07441","created_at":"2026-05-17T23:39:00Z"},{"alias_kind":"arxiv_version","alias_value":"2602.07441v2","created_at":"2026-05-17T23:39:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.07441","created_at":"2026-05-17T23:39:00Z"},{"alias_kind":"pith_short_12","alias_value":"LLYFYPQ7KKAX","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LLYFYPQ7KKAXCZ5C","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LLYFYPQ7","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:LLYFYPQ7KKAXCZ5CC5ET4AMJQ5","target":"record","payload":{"canonical_record":{"source":{"id":"2602.07441","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-07T08:44:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"11ceda1decf65b89cb4513c31144ecd8ad1132fdbf83a6e4a8002b631649229e","abstract_canon_sha256":"dd379b69b869ef4149d7ec98a96f6edfb06f271348c99998106e1a120b549e75"},"schema_version":"1.0"},"canonical_sha256":"5af05c3e1f52817167a217493e018987643dfe7037bdd563807d036431926018","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:00.026965Z","signature_b64":"Vg+Kd9qWxY32oyaRVXD2CfocUR6iYqKTWKhJzoO7U0tG+AFOgbIt7DaiX5C2mkfpZ2VjA1rv3/L+GscDf6sLAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5af05c3e1f52817167a217493e018987643dfe7037bdd563807d036431926018","last_reissued_at":"2026-05-17T23:39:00.026294Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:00.026294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.07441","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ivLIQqPDWyV0EPIP8HeZDkvKSdGH1mS3zzGUm656hHT5FxXLR+FqMdjGebCwzx32Be6tDKr/qLLS1BCUMiEKDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-04T15:07:13.276759Z"},"content_sha256":"35cbf042ce782890c59f19ebefa53cf65acc03bb94b0684762db451c247aefec","schema_version":"1.0","event_id":"sha256:35cbf042ce782890c59f19ebefa53cf65acc03bb94b0684762db451c247aefec"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:LLYFYPQ7KKAXCZ5CC5ET4AMJQ5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Proximal Action Replacement for Behavior Cloning Actor-Critic in Offline Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Proximal action replacement overcomes the imitation ceiling in BC-regularized actor-critic by substituting suboptimal dataset actions with value-guided improvements.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jianshu Zhang, Jinzong Dong, Nanyang Ye, Qinying Gu, Wei Huang, Xinzhe Yuan, Zhaohui Jiang, Zhuo Chen","submitted_at":"2026-02-07T08:44:27Z","abstract_excerpt":"Offline reinforcement learning (RL), which optimizes policies using a previously collected static dataset, is an important branch of RL. A popular and promising approach is to regularize actor-critic methods with behavior cloning (BC), which quickly yields realistic policies and mitigates bias from out-of-distribution actions, but it can impose an often-overlooked performance ceiling: when dataset actions are suboptimal, indiscriminate imitation structurally prevents the actor from fully exploiting better actions suggested by the value function, especially in later training when imitation is a"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"PAR consistently improves performance across offline RL benchmarks and approaches state-of-the-art results simply by being combined with the basic TD3+BC.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That actions generated by the stable target policy, guided by local ascent of the action-value function and bounded by value uncertainty, can be substituted without destabilizing training or introducing new bias when dataset actions are suboptimal.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Proximal action replacement breaks the imitation ceiling in BC-regularized offline RL actor-critic by substituting suboptimal dataset actions with value-guided improvements from a stable target policy.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Proximal action replacement overcomes the imitation ceiling in BC-regularized actor-critic by substituting suboptimal dataset actions with value-guided improvements.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"1bc0a1fabe3a824f2cbee31b02a28a8816be6ac66e3e322e509f4b137528e168"},"source":{"id":"2602.07441","kind":"arxiv","version":2},"verdict":{"id":"6c7949de-5314-4c6a-a646-019cf8846fd1","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T06:35:10.381452Z","strongest_claim":"PAR consistently improves performance across offline RL benchmarks and approaches state-of-the-art results simply by being combined with the basic TD3+BC.","one_line_summary":"Proximal action replacement breaks the imitation ceiling in BC-regularized offline RL actor-critic by substituting suboptimal dataset actions with value-guided improvements from a stable target policy.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That actions generated by the stable target policy, guided by local ascent of the action-value function and bounded by value uncertainty, can be substituted without destabilizing training or introducing new bias when dataset actions are suboptimal.","pith_extraction_headline":"Proximal action replacement overcomes the imitation ceiling in BC-regularized actor-critic by substituting suboptimal dataset actions with value-guided improvements."},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"b3849650427f66a860cf10c7b6fc96a1e819ae0ca1a3067c74188460cb719665"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"6c7949de-5314-4c6a-a646-019cf8846fd1"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"a2LwFD8CPFITee4g9u1PqfmtICpSv/j2D/8x9GJGYXeUBojuSxqjUKp3plB6JTXGQD9kTc6LK9VbX+yi+CsYBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-04T15:07:13.277286Z"},"content_sha256":"d448b925ee0578524a77ebcbb4c7a0c594873ed56af74ce586bc0c7ad25c6b17","schema_version":"1.0","event_id":"sha256:d448b925ee0578524a77ebcbb4c7a0c594873ed56af74ce586bc0c7ad25c6b17"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5/bundle.json","state_url":"https://pith.science/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-04T15:07:13Z","links":{"resolver":"https://pith.science/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5","bundle":"https://pith.science/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5/bundle.json","state":"https://pith.science/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LLYFYPQ7KKAXCZ5CC5ET4AMJQ5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:LLYFYPQ7KKAXCZ5CC5ET4AMJQ5","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":"dd379b69b869ef4149d7ec98a96f6edfb06f271348c99998106e1a120b549e75","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-07T08:44:27Z","title_canon_sha256":"11ceda1decf65b89cb4513c31144ecd8ad1132fdbf83a6e4a8002b631649229e"},"schema_version":"1.0","source":{"id":"2602.07441","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.07441","created_at":"2026-05-17T23:39:00Z"},{"alias_kind":"arxiv_version","alias_value":"2602.07441v2","created_at":"2026-05-17T23:39:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.07441","created_at":"2026-05-17T23:39:00Z"},{"alias_kind":"pith_short_12","alias_value":"LLYFYPQ7KKAX","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LLYFYPQ7KKAXCZ5C","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LLYFYPQ7","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:d448b925ee0578524a77ebcbb4c7a0c594873ed56af74ce586bc0c7ad25c6b17","target":"graph","created_at":"2026-05-17T23:39:00Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"PAR consistently improves performance across offline RL benchmarks and approaches state-of-the-art results simply by being combined with the basic TD3+BC."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That actions generated by the stable target policy, guided by local ascent of the action-value function and bounded by value uncertainty, can be substituted without destabilizing training or introducing new bias when dataset actions are suboptimal."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Proximal action replacement breaks the imitation ceiling in BC-regularized offline RL actor-critic by substituting suboptimal dataset actions with value-guided improvements from a stable target policy."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Proximal action replacement overcomes the imitation ceiling in BC-regularized actor-critic by substituting suboptimal dataset actions with value-guided improvements."}],"snapshot_sha256":"1bc0a1fabe3a824f2cbee31b02a28a8816be6ac66e3e322e509f4b137528e168"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"b3849650427f66a860cf10c7b6fc96a1e819ae0ca1a3067c74188460cb719665"},"paper":{"abstract_excerpt":"Offline reinforcement learning (RL), which optimizes policies using a previously collected static dataset, is an important branch of RL. A popular and promising approach is to regularize actor-critic methods with behavior cloning (BC), which quickly yields realistic policies and mitigates bias from out-of-distribution actions, but it can impose an often-overlooked performance ceiling: when dataset actions are suboptimal, indiscriminate imitation structurally prevents the actor from fully exploiting better actions suggested by the value function, especially in later training when imitation is a","authors_text":"Jianshu Zhang, Jinzong Dong, Nanyang Ye, Qinying Gu, Wei Huang, Xinzhe Yuan, Zhaohui Jiang, Zhuo Chen","cross_cats":["cs.AI"],"headline":"Proximal action replacement overcomes the imitation ceiling in BC-regularized actor-critic by substituting suboptimal dataset actions with value-guided improvements.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-07T08:44:27Z","title":"Proximal Action Replacement for Behavior Cloning Actor-Critic in Offline Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.07441","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-16T06:35:10.381452Z","id":"6c7949de-5314-4c6a-a646-019cf8846fd1","model_set":{"reader":"grok-4.3"},"one_line_summary":"Proximal action replacement breaks the imitation ceiling in BC-regularized offline RL actor-critic by substituting suboptimal dataset actions with value-guided improvements from a stable target policy.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Proximal action replacement overcomes the imitation ceiling in BC-regularized actor-critic by substituting suboptimal dataset actions with value-guided improvements.","strongest_claim":"PAR consistently improves performance across offline RL benchmarks and approaches state-of-the-art results simply by being combined with the basic TD3+BC.","weakest_assumption":"That actions generated by the stable target policy, guided by local ascent of the action-value function and bounded by value uncertainty, can be substituted without destabilizing training or introducing new bias when dataset actions are suboptimal."}},"verdict_id":"6c7949de-5314-4c6a-a646-019cf8846fd1"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:35cbf042ce782890c59f19ebefa53cf65acc03bb94b0684762db451c247aefec","target":"record","created_at":"2026-05-17T23:39:00Z","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":"dd379b69b869ef4149d7ec98a96f6edfb06f271348c99998106e1a120b549e75","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-07T08:44:27Z","title_canon_sha256":"11ceda1decf65b89cb4513c31144ecd8ad1132fdbf83a6e4a8002b631649229e"},"schema_version":"1.0","source":{"id":"2602.07441","kind":"arxiv","version":2}},"canonical_sha256":"5af05c3e1f52817167a217493e018987643dfe7037bdd563807d036431926018","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5af05c3e1f52817167a217493e018987643dfe7037bdd563807d036431926018","first_computed_at":"2026-05-17T23:39:00.026294Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:00.026294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Vg+Kd9qWxY32oyaRVXD2CfocUR6iYqKTWKhJzoO7U0tG+AFOgbIt7DaiX5C2mkfpZ2VjA1rv3/L+GscDf6sLAg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:00.026965Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.07441","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:35cbf042ce782890c59f19ebefa53cf65acc03bb94b0684762db451c247aefec","sha256:d448b925ee0578524a77ebcbb4c7a0c594873ed56af74ce586bc0c7ad25c6b17"],"state_sha256":"97db272826964c1087c15a871d42daacc17d974f977d0fddf8bff50bcf2e9297"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OWgjwFuDQPsoSYVvWCnEWep6QO6k3sqoyBkaRm5xeDA48dKNCkk88RsYbvQVa+4VAeHyxou6XmaXdoBR1ExnBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-04T15:07:13.280345Z","bundle_sha256":"dd920ac083a232282bcc076964c89144240eabefd4a93ab2be7832e33f760a8e"}}