{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:XQMZ2XJZBU7CTUBY2YSA25UM2H","short_pith_number":"pith:XQMZ2XJZ","canonical_record":{"source":{"id":"2106.06630","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-06-11T22:41:53Z","cross_cats_sorted":[],"title_canon_sha256":"ae9dc094ff4f2cfb3ebd3ec5948ed1ea535e863375da5a29b0969494baf6a0bc","abstract_canon_sha256":"453c1546a8a2920109a542ad674be591b4c5f592f8e5fda152b3781e452186e4"},"schema_version":"1.0"},"canonical_sha256":"bc199d5d390d3e29d038d6240d768cd1fe589fe7081ce7c2ae656244ec46d03f","source":{"kind":"arxiv","id":"2106.06630","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.06630","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"arxiv_version","alias_value":"2106.06630v1","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.06630","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"pith_short_12","alias_value":"XQMZ2XJZBU7C","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"pith_short_16","alias_value":"XQMZ2XJZBU7CTUBY","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"pith_short_8","alias_value":"XQMZ2XJZ","created_at":"2026-07-05T02:48:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:XQMZ2XJZBU7CTUBY2YSA25UM2H","target":"record","payload":{"canonical_record":{"source":{"id":"2106.06630","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-06-11T22:41:53Z","cross_cats_sorted":[],"title_canon_sha256":"ae9dc094ff4f2cfb3ebd3ec5948ed1ea535e863375da5a29b0969494baf6a0bc","abstract_canon_sha256":"453c1546a8a2920109a542ad674be591b4c5f592f8e5fda152b3781e452186e4"},"schema_version":"1.0"},"canonical_sha256":"bc199d5d390d3e29d038d6240d768cd1fe589fe7081ce7c2ae656244ec46d03f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:48:39.754597Z","signature_b64":"AA2jBRK55xGNgcWGC9z/tfKKWLMmxrOIvYvie8BJZDO9ypbGzETey/VMj0333N6ix6W1mp+1wDknPkD2J8SrAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bc199d5d390d3e29d038d6240d768cd1fe589fe7081ce7c2ae656244ec46d03f","last_reissued_at":"2026-07-05T02:48:39.754120Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:48:39.754120Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2106.06630","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T02:48:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wvkooRUEagjTrp9RuKn24iFWCMl7BcaybzJ7UmgaOcYkMbeTENJ+9yCWBGuPwhpTSVLGJm0bvUIaDXKx5J01Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T19:43:26.419928Z"},"content_sha256":"1cafdc7fd9fd7a4032b3fed1ff0991ff9ec48c765cdf4fa2ca206648db01a7fe","schema_version":"1.0","event_id":"sha256:1cafdc7fd9fd7a4032b3fed1ff0991ff9ec48c765cdf4fa2ca206648db01a7fe"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:XQMZ2XJZBU7CTUBY2YSA25UM2H","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Corruption-Robust Offline Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jerry Zhu, Wen Sun, Xuezhou Zhang, Yiding Chen","submitted_at":"2021-06-11T22:41:53Z","abstract_excerpt":"We study the adversarial robustness in offline reinforcement learning. Given a batch dataset consisting of tuples $(s, a, r, s')$, an adversary is allowed to arbitrarily modify $\\epsilon$ fraction of the tuples. From the corrupted dataset the learner aims to robustly identify a near-optimal policy. We first show that a worst-case $\\Omega(d\\epsilon)$ optimality gap is unavoidable in linear MDP of dimension $d$, even if the adversary only corrupts the reward element in a tuple. This contrasts with dimension-free results in robust supervised learning and best-known lower-bound in the online RL se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.06630","kind":"arxiv","version":1},"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/2106.06630/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":"2026-07-05T02:48:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RpUU/LnVvn8Jkk0nGOrAA8pJFMGx0pJpa00oN8lORDrc78Sri2/Ee5NNaAcrheOur277yNdply+QMKRxL2LgAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T19:43:26.420692Z"},"content_sha256":"e7930d8f61e9c1518ea4ed1b02f17df2c9a88d9abc7f8320b62e58c63413e0b5","schema_version":"1.0","event_id":"sha256:e7930d8f61e9c1518ea4ed1b02f17df2c9a88d9abc7f8320b62e58c63413e0b5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H/bundle.json","state_url":"https://pith.science/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H/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":"2026-08-08T19:43:26Z","links":{"resolver":"https://pith.science/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H","bundle":"https://pith.science/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H/bundle.json","state":"https://pith.science/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XQMZ2XJZBU7CTUBY2YSA25UM2H/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:XQMZ2XJZBU7CTUBY2YSA25UM2H","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":"453c1546a8a2920109a542ad674be591b4c5f592f8e5fda152b3781e452186e4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-06-11T22:41:53Z","title_canon_sha256":"ae9dc094ff4f2cfb3ebd3ec5948ed1ea535e863375da5a29b0969494baf6a0bc"},"schema_version":"1.0","source":{"id":"2106.06630","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.06630","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"arxiv_version","alias_value":"2106.06630v1","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.06630","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"pith_short_12","alias_value":"XQMZ2XJZBU7C","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"pith_short_16","alias_value":"XQMZ2XJZBU7CTUBY","created_at":"2026-07-05T02:48:39Z"},{"alias_kind":"pith_short_8","alias_value":"XQMZ2XJZ","created_at":"2026-07-05T02:48:39Z"}],"graph_snapshots":[{"event_id":"sha256:e7930d8f61e9c1518ea4ed1b02f17df2c9a88d9abc7f8320b62e58c63413e0b5","target":"graph","created_at":"2026-07-05T02:48:39Z","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/2106.06630/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We study the adversarial robustness in offline reinforcement learning. Given a batch dataset consisting of tuples $(s, a, r, s')$, an adversary is allowed to arbitrarily modify $\\epsilon$ fraction of the tuples. From the corrupted dataset the learner aims to robustly identify a near-optimal policy. We first show that a worst-case $\\Omega(d\\epsilon)$ optimality gap is unavoidable in linear MDP of dimension $d$, even if the adversary only corrupts the reward element in a tuple. This contrasts with dimension-free results in robust supervised learning and best-known lower-bound in the online RL se","authors_text":"Jerry Zhu, Wen Sun, Xuezhou Zhang, Yiding Chen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-06-11T22:41:53Z","title":"Corruption-Robust Offline Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.06630","kind":"arxiv","version":1},"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":"sha256:1cafdc7fd9fd7a4032b3fed1ff0991ff9ec48c765cdf4fa2ca206648db01a7fe","target":"record","created_at":"2026-07-05T02:48:39Z","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":"453c1546a8a2920109a542ad674be591b4c5f592f8e5fda152b3781e452186e4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-06-11T22:41:53Z","title_canon_sha256":"ae9dc094ff4f2cfb3ebd3ec5948ed1ea535e863375da5a29b0969494baf6a0bc"},"schema_version":"1.0","source":{"id":"2106.06630","kind":"arxiv","version":1}},"canonical_sha256":"bc199d5d390d3e29d038d6240d768cd1fe589fe7081ce7c2ae656244ec46d03f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bc199d5d390d3e29d038d6240d768cd1fe589fe7081ce7c2ae656244ec46d03f","first_computed_at":"2026-07-05T02:48:39.754120Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:48:39.754120Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"AA2jBRK55xGNgcWGC9z/tfKKWLMmxrOIvYvie8BJZDO9ypbGzETey/VMj0333N6ix6W1mp+1wDknPkD2J8SrAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T02:48:39.754597Z","signed_message":"canonical_sha256_bytes"},"source_id":"2106.06630","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1cafdc7fd9fd7a4032b3fed1ff0991ff9ec48c765cdf4fa2ca206648db01a7fe","sha256:e7930d8f61e9c1518ea4ed1b02f17df2c9a88d9abc7f8320b62e58c63413e0b5"],"state_sha256":"4cd3149caf00dd8d27baeb18b7324dd14897053c4e4d30f0eb21288239e5e818"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"21EpTtzSklzJehXeQW0AHpamKz8lKkIbZ673pvWoDkmrrlDSj/PlNSXkMSU6EwsMFWzkK0GOepYk2GmS8yLPDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T19:43:26.425169Z","bundle_sha256":"b051ba34acd3ea9b66a823d39e2a0ea0a51c692e06a908278aa62861929b8f74"}}