{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:3YRDTMD44J2FPCWYPMEW5IPT7D","short_pith_number":"pith:3YRDTMD4","canonical_record":{"source":{"id":"2102.00479","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-01-31T16:17:56Z","cross_cats_sorted":["cs.AI","math.OC","stat.ML"],"title_canon_sha256":"0cfe26f18752674f2444fdd1735b9db74e6ae6e574f9cbb5e213b2bfe841e5a3","abstract_canon_sha256":"ea5d6221122bdb9393fa688ccbe3f9e081eb5d6ef950d4d41868e4c90047590b"},"schema_version":"1.0"},"canonical_sha256":"de2239b07ce274578ad87b096ea1f3f8c4311cfffd3e1fc7431e6b9ff7a6e8e1","source":{"kind":"arxiv","id":"2102.00479","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2102.00479","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"arxiv_version","alias_value":"2102.00479v2","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.00479","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"pith_short_12","alias_value":"3YRDTMD44J2F","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"pith_short_16","alias_value":"3YRDTMD44J2FPCWY","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"pith_short_8","alias_value":"3YRDTMD4","created_at":"2026-07-05T06:29:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:3YRDTMD44J2FPCWYPMEW5IPT7D","target":"record","payload":{"canonical_record":{"source":{"id":"2102.00479","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-01-31T16:17:56Z","cross_cats_sorted":["cs.AI","math.OC","stat.ML"],"title_canon_sha256":"0cfe26f18752674f2444fdd1735b9db74e6ae6e574f9cbb5e213b2bfe841e5a3","abstract_canon_sha256":"ea5d6221122bdb9393fa688ccbe3f9e081eb5d6ef950d4d41868e4c90047590b"},"schema_version":"1.0"},"canonical_sha256":"de2239b07ce274578ad87b096ea1f3f8c4311cfffd3e1fc7431e6b9ff7a6e8e1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:29:59.609015Z","signature_b64":"Q5wp2PDs3in/mGnixoztPzmxQoM6VmHlBiI01noYiX/fs984TXvK9HYX+wgk83Sc8oNWLIm0OW83wOnuH022BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"de2239b07ce274578ad87b096ea1f3f8c4311cfffd3e1fc7431e6b9ff7a6e8e1","last_reissued_at":"2026-07-05T06:29:59.608548Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:29:59.608548Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2102.00479","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":"2026-07-05T06:29:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"y3GqQ4xYbFK+rrm0BmfyV9BPGMsTyB1JoH5rdQBisDuZAnTWQ/F2EAQZ/0e9vSCVG1Gc4/k8O7l2Ws9JZCO5Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T16:01:01.470007Z"},"content_sha256":"c5c131a0d70060d273f109a07810847af553c5f825660ac67a521c9e6baac63c","schema_version":"1.0","event_id":"sha256:c5c131a0d70060d273f109a07810847af553c5f825660ac67a521c9e6baac63c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:3YRDTMD44J2FPCWYPMEW5IPT7D","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fast Rates for the Regret of Offline Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Masatoshi Uehara, Nathan Kallus, Yichun Hu","submitted_at":"2021-01-31T16:17:56Z","abstract_excerpt":"We study the regret of reinforcement learning from offline data generated by a fixed behavior policy in an infinite-horizon discounted Markov decision process (MDP). While existing analyses of common approaches, such as fitted $Q$-iteration (FQI), suggest a $O(1/\\sqrt{n})$ convergence for regret, empirical behavior exhibits \\emph{much} faster convergence. In this paper, we present a finer regret analysis that exactly characterizes this phenomenon by providing fast rates for the regret convergence. First, we show that given any estimate for the optimal quality function $Q^*$, the regret of the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.00479","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2102.00479/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T06:29:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FfV2abyD6RkGyRaytqkRcqAAihcNpIZVCM9bJjvGufDKr280da//4ByBv7s5aB3PP+GOe/ZZ4mOCyNGlwsPXAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T16:01:01.470425Z"},"content_sha256":"c08b4e52249dbd5acda16ca1bf1066dce64abd35980a6b8c917f39f5b85fb6b1","schema_version":"1.0","event_id":"sha256:c08b4e52249dbd5acda16ca1bf1066dce64abd35980a6b8c917f39f5b85fb6b1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3YRDTMD44J2FPCWYPMEW5IPT7D/bundle.json","state_url":"https://pith.science/pith/3YRDTMD44J2FPCWYPMEW5IPT7D/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3YRDTMD44J2FPCWYPMEW5IPT7D/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":"2026-08-18T16:01:01Z","links":{"resolver":"https://pith.science/pith/3YRDTMD44J2FPCWYPMEW5IPT7D","bundle":"https://pith.science/pith/3YRDTMD44J2FPCWYPMEW5IPT7D/bundle.json","state":"https://pith.science/pith/3YRDTMD44J2FPCWYPMEW5IPT7D/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3YRDTMD44J2FPCWYPMEW5IPT7D/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:3YRDTMD44J2FPCWYPMEW5IPT7D","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":"ea5d6221122bdb9393fa688ccbe3f9e081eb5d6ef950d4d41868e4c90047590b","cross_cats_sorted":["cs.AI","math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-01-31T16:17:56Z","title_canon_sha256":"0cfe26f18752674f2444fdd1735b9db74e6ae6e574f9cbb5e213b2bfe841e5a3"},"schema_version":"1.0","source":{"id":"2102.00479","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2102.00479","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"arxiv_version","alias_value":"2102.00479v2","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.00479","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"pith_short_12","alias_value":"3YRDTMD44J2F","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"pith_short_16","alias_value":"3YRDTMD44J2FPCWY","created_at":"2026-07-05T06:29:59Z"},{"alias_kind":"pith_short_8","alias_value":"3YRDTMD4","created_at":"2026-07-05T06:29:59Z"}],"graph_snapshots":[{"event_id":"sha256:c08b4e52249dbd5acda16ca1bf1066dce64abd35980a6b8c917f39f5b85fb6b1","target":"graph","created_at":"2026-07-05T06:29:59Z","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/2102.00479/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We study the regret of reinforcement learning from offline data generated by a fixed behavior policy in an infinite-horizon discounted Markov decision process (MDP). While existing analyses of common approaches, such as fitted $Q$-iteration (FQI), suggest a $O(1/\\sqrt{n})$ convergence for regret, empirical behavior exhibits \\emph{much} faster convergence. In this paper, we present a finer regret analysis that exactly characterizes this phenomenon by providing fast rates for the regret convergence. First, we show that given any estimate for the optimal quality function $Q^*$, the regret of the ","authors_text":"Masatoshi Uehara, Nathan Kallus, Yichun Hu","cross_cats":["cs.AI","math.OC","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-01-31T16:17:56Z","title":"Fast Rates for the Regret of Offline Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.00479","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:c5c131a0d70060d273f109a07810847af553c5f825660ac67a521c9e6baac63c","target":"record","created_at":"2026-07-05T06:29:59Z","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":"ea5d6221122bdb9393fa688ccbe3f9e081eb5d6ef950d4d41868e4c90047590b","cross_cats_sorted":["cs.AI","math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-01-31T16:17:56Z","title_canon_sha256":"0cfe26f18752674f2444fdd1735b9db74e6ae6e574f9cbb5e213b2bfe841e5a3"},"schema_version":"1.0","source":{"id":"2102.00479","kind":"arxiv","version":2}},"canonical_sha256":"de2239b07ce274578ad87b096ea1f3f8c4311cfffd3e1fc7431e6b9ff7a6e8e1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"de2239b07ce274578ad87b096ea1f3f8c4311cfffd3e1fc7431e6b9ff7a6e8e1","first_computed_at":"2026-07-05T06:29:59.608548Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:29:59.608548Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Q5wp2PDs3in/mGnixoztPzmxQoM6VmHlBiI01noYiX/fs984TXvK9HYX+wgk83Sc8oNWLIm0OW83wOnuH022BQ==","signature_status":"signed_v1","signed_at":"2026-07-05T06:29:59.609015Z","signed_message":"canonical_sha256_bytes"},"source_id":"2102.00479","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c5c131a0d70060d273f109a07810847af553c5f825660ac67a521c9e6baac63c","sha256:c08b4e52249dbd5acda16ca1bf1066dce64abd35980a6b8c917f39f5b85fb6b1"],"state_sha256":"ec9156e666052077e506ccd2888adf877122071b4142046fb7f23dc0d66b0ade"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SNQ6dktOfaPh3lXrMc5T+AekfknFPsTQUhmbXTWCBE7ryu4QybAkg0nAh3qQjtTRr2YfrcJ9j8uNTcvgL6yjDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T16:01:01.485928Z","bundle_sha256":"46a2a23e763c6a8656dd9f3db12f9c53f0b049a0cf273d8552e613d22f873a9c"}}