{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XD27MZ74LOODBVCMJDOHK63S33","short_pith_number":"pith:XD27MZ74","schema_version":"1.0","canonical_sha256":"b8f5f667fc5b9c30d44c48dc757b72deed88ef3223cf1c15cd2b55eff588b59c","source":{"kind":"arxiv","id":"2405.14114","version":2},"attestation_state":"computed","paper":{"title":"Offline Reinforcement Learning from Datasets with Structured Non-Stationarity","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Johannes Ackermann, Masashi Sugiyama, Takayuki Osa","submitted_at":"2024-05-23T02:41:36Z","abstract_excerpt":"Current Reinforcement Learning (RL) is often limited by the large amount of data needed to learn a successful policy. Offline RL aims to solve this issue by using transitions collected by a different behavior policy. We address a novel Offline RL problem setting in which, while collecting the dataset, the transition and reward functions gradually change between episodes but stay constant within each episode. We propose a method based on Contrastive Predictive Coding that identifies this non-stationarity in the offline dataset, accounts for it when training a policy, and predicts it during eval"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2405.14114","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-23T02:41:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bc91669e851a9fd128e4844f4b1eb9e82c87c61a4a894112028c291114d6a602","abstract_canon_sha256":"30880fa73b09851d6ce3dd8d92b06b63ee602e021cc58256ec0e625e49320c37"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:06.173360Z","signature_b64":"2vPPTNYPzHmiYTXEA4DtxuozzqyRtrDKFYdsa8gOA08mz4kKmkz/hCqMId2PDHEoCMY/q+CY1ETXLMGXpFqGBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b8f5f667fc5b9c30d44c48dc757b72deed88ef3223cf1c15cd2b55eff588b59c","last_reissued_at":"2026-07-05T08:24:06.172820Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:06.172820Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Offline Reinforcement Learning from Datasets with Structured Non-Stationarity","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Johannes Ackermann, Masashi Sugiyama, Takayuki Osa","submitted_at":"2024-05-23T02:41:36Z","abstract_excerpt":"Current Reinforcement Learning (RL) is often limited by the large amount of data needed to learn a successful policy. Offline RL aims to solve this issue by using transitions collected by a different behavior policy. We address a novel Offline RL problem setting in which, while collecting the dataset, the transition and reward functions gradually change between episodes but stay constant within each episode. We propose a method based on Contrastive Predictive Coding that identifies this non-stationarity in the offline dataset, accounts for it when training a policy, and predicts it during eval"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.14114","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/2405.14114/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2405.14114","created_at":"2026-07-05T08:24:06.172878+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.14114v2","created_at":"2026-07-05T08:24:06.172878+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.14114","created_at":"2026-07-05T08:24:06.172878+00:00"},{"alias_kind":"pith_short_12","alias_value":"XD27MZ74LOOD","created_at":"2026-07-05T08:24:06.172878+00:00"},{"alias_kind":"pith_short_16","alias_value":"XD27MZ74LOODBVCM","created_at":"2026-07-05T08:24:06.172878+00:00"},{"alias_kind":"pith_short_8","alias_value":"XD27MZ74","created_at":"2026-07-05T08:24:06.172878+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33","json":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33.json","graph_json":"https://pith.science/api/pith-number/XD27MZ74LOODBVCMJDOHK63S33/graph.json","events_json":"https://pith.science/api/pith-number/XD27MZ74LOODBVCMJDOHK63S33/events.json","paper":"https://pith.science/paper/XD27MZ74"},"agent_actions":{"view_html":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33","download_json":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33.json","view_paper":"https://pith.science/paper/XD27MZ74","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.14114&json=true","fetch_graph":"https://pith.science/api/pith-number/XD27MZ74LOODBVCMJDOHK63S33/graph.json","fetch_events":"https://pith.science/api/pith-number/XD27MZ74LOODBVCMJDOHK63S33/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33/action/storage_attestation","attest_author":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33/action/author_attestation","sign_citation":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33/action/citation_signature","submit_replication":"https://pith.science/pith/XD27MZ74LOODBVCMJDOHK63S33/action/replication_record"}},"created_at":"2026-07-05T08:24:06.172878+00:00","updated_at":"2026-07-05T08:24:06.172878+00:00"}