{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:PN42TRUCPLLYHUK6KTUPHKUU4C","short_pith_number":"pith:PN42TRUC","canonical_record":{"source":{"id":"2111.14552","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-29T14:30:26Z","cross_cats_sorted":[],"title_canon_sha256":"0d0c0a44304c8f91c16a60bd7a5805ff65b62a0424529cc1661e2428610c3d8e","abstract_canon_sha256":"607abda45fc84398dbbbb1164caef13976a1bf213cafb9d159be055311f16e56"},"schema_version":"1.0"},"canonical_sha256":"7b79a9c6827ad783d15e54e8f3aa94e0a06f8272d14d4f71d7eed784417503a2","source":{"kind":"arxiv","id":"2111.14552","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2111.14552","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"arxiv_version","alias_value":"2111.14552v2","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.14552","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"pith_short_12","alias_value":"PN42TRUCPLLY","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"pith_short_16","alias_value":"PN42TRUCPLLYHUK6","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"pith_short_8","alias_value":"PN42TRUC","created_at":"2026-07-05T05:05:01Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:PN42TRUCPLLYHUK6KTUPHKUU4C","target":"record","payload":{"canonical_record":{"source":{"id":"2111.14552","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-29T14:30:26Z","cross_cats_sorted":[],"title_canon_sha256":"0d0c0a44304c8f91c16a60bd7a5805ff65b62a0424529cc1661e2428610c3d8e","abstract_canon_sha256":"607abda45fc84398dbbbb1164caef13976a1bf213cafb9d159be055311f16e56"},"schema_version":"1.0"},"canonical_sha256":"7b79a9c6827ad783d15e54e8f3aa94e0a06f8272d14d4f71d7eed784417503a2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:05:01.879116Z","signature_b64":"vG2B6fzkzfaDoKARNGHJrNZcA7sYRbu7VTqOWv/JYC+OGYnwuWtxv4l5n/NCwAZin3MkuRt96Sy/nHzBmQZ5Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7b79a9c6827ad783d15e54e8f3aa94e0a06f8272d14d4f71d7eed784417503a2","last_reissued_at":"2026-07-05T05:05:01.878612Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:05:01.878612Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2111.14552","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":"2026-07-05T05:05:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8TAaWEVwRMmmd3PhMS6WnHgITKZ5SmnPje1J6psdNZv/HD938RMplmVnfnl6yQUzibpTiLE6YtIx+bzSWj9gCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T05:50:16.210969Z"},"content_sha256":"706158e4ec88235ce0aecf8d090c71e63d12ed1b314921f78304db8180d60f1b","schema_version":"1.0","event_id":"sha256:706158e4ec88235ce0aecf8d090c71e63d12ed1b314921f78304db8180d60f1b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:PN42TRUCPLLYHUK6KTUPHKUU4C","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Robust On-Policy Sampling for Data-Efficient Policy Evaluation in Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Duohan Zhang, Josiah P. Hanna, Lukas Sch\\\"afer, Rujie Zhong, Stefano V. Albrecht","submitted_at":"2021-11-29T14:30:26Z","abstract_excerpt":"Reinforcement learning (RL) algorithms are often categorized as either on-policy or off-policy depending on whether they use data from a target policy of interest or from a different behavior policy. In this paper, we study a subtle distinction between on-policy data and on-policy sampling in the context of the RL sub-problem of policy evaluation. We observe that on-policy sampling may fail to match the expected distribution of on-policy data after observing only a finite number of trajectories and this failure hinders data-efficient policy evaluation. Towards improved data-efficiency, we show"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.14552","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/2111.14552/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":"2026-07-05T05:05:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7Rcyty9CeqUnFXGzKG3kMOEcIcaVIEUS82Qry4RkafgMgYdbpWHFcnM45kYg+CwG9rNm/P8PISQTEuKkJeH1Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T05:50:16.211474Z"},"content_sha256":"f6dc7a0086347673b883130b5773f9ebbe5b3c55cfe744364261bdd7c10177d9","schema_version":"1.0","event_id":"sha256:f6dc7a0086347673b883130b5773f9ebbe5b3c55cfe744364261bdd7c10177d9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PN42TRUCPLLYHUK6KTUPHKUU4C/bundle.json","state_url":"https://pith.science/pith/PN42TRUCPLLYHUK6KTUPHKUU4C/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PN42TRUCPLLYHUK6KTUPHKUU4C/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":"2026-08-11T05:50:16Z","links":{"resolver":"https://pith.science/pith/PN42TRUCPLLYHUK6KTUPHKUU4C","bundle":"https://pith.science/pith/PN42TRUCPLLYHUK6KTUPHKUU4C/bundle.json","state":"https://pith.science/pith/PN42TRUCPLLYHUK6KTUPHKUU4C/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PN42TRUCPLLYHUK6KTUPHKUU4C/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:PN42TRUCPLLYHUK6KTUPHKUU4C","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":"607abda45fc84398dbbbb1164caef13976a1bf213cafb9d159be055311f16e56","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-29T14:30:26Z","title_canon_sha256":"0d0c0a44304c8f91c16a60bd7a5805ff65b62a0424529cc1661e2428610c3d8e"},"schema_version":"1.0","source":{"id":"2111.14552","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2111.14552","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"arxiv_version","alias_value":"2111.14552v2","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.14552","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"pith_short_12","alias_value":"PN42TRUCPLLY","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"pith_short_16","alias_value":"PN42TRUCPLLYHUK6","created_at":"2026-07-05T05:05:01Z"},{"alias_kind":"pith_short_8","alias_value":"PN42TRUC","created_at":"2026-07-05T05:05:01Z"}],"graph_snapshots":[{"event_id":"sha256:f6dc7a0086347673b883130b5773f9ebbe5b3c55cfe744364261bdd7c10177d9","target":"graph","created_at":"2026-07-05T05:05:01Z","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/2111.14552/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning (RL) algorithms are often categorized as either on-policy or off-policy depending on whether they use data from a target policy of interest or from a different behavior policy. In this paper, we study a subtle distinction between on-policy data and on-policy sampling in the context of the RL sub-problem of policy evaluation. We observe that on-policy sampling may fail to match the expected distribution of on-policy data after observing only a finite number of trajectories and this failure hinders data-efficient policy evaluation. Towards improved data-efficiency, we show","authors_text":"Duohan Zhang, Josiah P. Hanna, Lukas Sch\\\"afer, Rujie Zhong, Stefano V. Albrecht","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-29T14:30:26Z","title":"Robust On-Policy Sampling for Data-Efficient Policy Evaluation in Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.14552","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":"sha256:706158e4ec88235ce0aecf8d090c71e63d12ed1b314921f78304db8180d60f1b","target":"record","created_at":"2026-07-05T05:05:01Z","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":"607abda45fc84398dbbbb1164caef13976a1bf213cafb9d159be055311f16e56","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-29T14:30:26Z","title_canon_sha256":"0d0c0a44304c8f91c16a60bd7a5805ff65b62a0424529cc1661e2428610c3d8e"},"schema_version":"1.0","source":{"id":"2111.14552","kind":"arxiv","version":2}},"canonical_sha256":"7b79a9c6827ad783d15e54e8f3aa94e0a06f8272d14d4f71d7eed784417503a2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7b79a9c6827ad783d15e54e8f3aa94e0a06f8272d14d4f71d7eed784417503a2","first_computed_at":"2026-07-05T05:05:01.878612Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:05:01.878612Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vG2B6fzkzfaDoKARNGHJrNZcA7sYRbu7VTqOWv/JYC+OGYnwuWtxv4l5n/NCwAZin3MkuRt96Sy/nHzBmQZ5Dg==","signature_status":"signed_v1","signed_at":"2026-07-05T05:05:01.879116Z","signed_message":"canonical_sha256_bytes"},"source_id":"2111.14552","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:706158e4ec88235ce0aecf8d090c71e63d12ed1b314921f78304db8180d60f1b","sha256:f6dc7a0086347673b883130b5773f9ebbe5b3c55cfe744364261bdd7c10177d9"],"state_sha256":"4b44d5ebda46e848808cf6f2fa3930a3aa6d446a4a618fa01db417e919bddbcc"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3F32i7+InlvgXNhVroosDfefKNt3WijFi/jQEjznHsrDBl0xMGq3znNazLniVh+Us/fu640lbArpbVnIfooRBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-11T05:50:16.218055Z","bundle_sha256":"d98e296decc2459b82e4494fd030168d5b9e51ce669a39539a41295fedc0887f"}}