{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YZ7IIG55CZQ6GFBEH6P2UUJKK5","short_pith_number":"pith:YZ7IIG55","schema_version":"1.0","canonical_sha256":"c67e841bbd1661e314243f9faa512a5774cc4cc0fd0ecdecd4fb2eb7ede635d6","source":{"kind":"arxiv","id":"2407.14653","version":1},"attestation_state":"computed","paper":{"title":"OASIS: Conditional Distribution Shaping for Offline Safe Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ding Zhao, Haohong Lin, Shiqi Liu, Tingnan Zhang, Wenhao Ding, Wenhao Yu, Yihang Yao, Zhepeng Cen","submitted_at":"2024-07-19T20:15:00Z","abstract_excerpt":"Offline safe reinforcement learning (RL) aims to train a policy that satisfies constraints using a pre-collected dataset. Most current methods struggle with the mismatch between imperfect demonstrations and the desired safe and rewarding performance. In this paper, we introduce OASIS (cOnditionAl diStributIon Shaping), a new paradigm in offline safe RL designed to overcome these critical limitations. OASIS utilizes a conditional diffusion model to synthesize offline datasets, thus shaping the data distribution toward a beneficial target domain. Our approach makes compliance with safety constra"},"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":"2407.14653","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-19T20:15:00Z","cross_cats_sorted":[],"title_canon_sha256":"1b1c2b73dae31576855114e9ba41ef4169cc06671b0df595d3f1a9f3ac2130cf","abstract_canon_sha256":"89f250d53f4abc25708ddda9e80a2191bc4dc6ade87520bf844715690c98921d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:46:31.777394Z","signature_b64":"GrxCunRa9v9yUXpAQnIPg8ntHJ7rys/EFJ067zcOjtkrpFcZ7h77b5Bd6GBwFyDGZCGT70W9oTB9tiK73KkbCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c67e841bbd1661e314243f9faa512a5774cc4cc0fd0ecdecd4fb2eb7ede635d6","last_reissued_at":"2026-07-05T08:46:31.776962Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:46:31.776962Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OASIS: Conditional Distribution Shaping for Offline Safe Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ding Zhao, Haohong Lin, Shiqi Liu, Tingnan Zhang, Wenhao Ding, Wenhao Yu, Yihang Yao, Zhepeng Cen","submitted_at":"2024-07-19T20:15:00Z","abstract_excerpt":"Offline safe reinforcement learning (RL) aims to train a policy that satisfies constraints using a pre-collected dataset. Most current methods struggle with the mismatch between imperfect demonstrations and the desired safe and rewarding performance. In this paper, we introduce OASIS (cOnditionAl diStributIon Shaping), a new paradigm in offline safe RL designed to overcome these critical limitations. OASIS utilizes a conditional diffusion model to synthesize offline datasets, thus shaping the data distribution toward a beneficial target domain. Our approach makes compliance with safety constra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.14653","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/2407.14653/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":"2407.14653","created_at":"2026-07-05T08:46:31.777019+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.14653v1","created_at":"2026-07-05T08:46:31.777019+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.14653","created_at":"2026-07-05T08:46:31.777019+00:00"},{"alias_kind":"pith_short_12","alias_value":"YZ7IIG55CZQ6","created_at":"2026-07-05T08:46:31.777019+00:00"},{"alias_kind":"pith_short_16","alias_value":"YZ7IIG55CZQ6GFBE","created_at":"2026-07-05T08:46:31.777019+00:00"},{"alias_kind":"pith_short_8","alias_value":"YZ7IIG55","created_at":"2026-07-05T08:46:31.777019+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.08880","citing_title":"FAWAC: Feasibility Informed Advantage Weighted Regression for Persistent Safety in Offline Reinforcement Learning","ref_index":30,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5","json":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5.json","graph_json":"https://pith.science/api/pith-number/YZ7IIG55CZQ6GFBEH6P2UUJKK5/graph.json","events_json":"https://pith.science/api/pith-number/YZ7IIG55CZQ6GFBEH6P2UUJKK5/events.json","paper":"https://pith.science/paper/YZ7IIG55"},"agent_actions":{"view_html":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5","download_json":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5.json","view_paper":"https://pith.science/paper/YZ7IIG55","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.14653&json=true","fetch_graph":"https://pith.science/api/pith-number/YZ7IIG55CZQ6GFBEH6P2UUJKK5/graph.json","fetch_events":"https://pith.science/api/pith-number/YZ7IIG55CZQ6GFBEH6P2UUJKK5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5/action/storage_attestation","attest_author":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5/action/author_attestation","sign_citation":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5/action/citation_signature","submit_replication":"https://pith.science/pith/YZ7IIG55CZQ6GFBEH6P2UUJKK5/action/replication_record"}},"created_at":"2026-07-05T08:46:31.777019+00:00","updated_at":"2026-07-05T08:46:31.777019+00:00"}