{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:TA3XZ3GCSPPJWUZQGDDI7ZCWPX","short_pith_number":"pith:TA3XZ3GC","schema_version":"1.0","canonical_sha256":"98377cecc293de9b533030c68fe4567ddb4a659d730313c9753f35447256a421","source":{"kind":"arxiv","id":"1902.08708","version":2},"attestation_state":"computed","paper":{"title":"Distributionally Robust Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Elena Smirnova, Elvis Dohmatob, J\\'er\\'emie Mary","submitted_at":"2019-02-23T00:13:42Z","abstract_excerpt":"Real-world applications require RL algorithms to act safely. During learning process, it is likely that the agent executes sub-optimal actions that may lead to unsafe/poor states of the system. Exploration is particularly brittle in high-dimensional state/action space due to increased number of low-performing actions. In this work, we consider risk-averse exploration in approximate RL setting. To ensure safety during learning, we propose the distributionally robust policy iteration scheme that provides lower bound guarantee on state-values. Our approach induces a dynamic level of risk to preve"},"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":"1902.08708","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2019-02-23T00:13:42Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"96114f83ff82599a2b4b0aaf363e4086f4af6d54410c463c60770fdafe67061f","abstract_canon_sha256":"599bd5b68dfb0a2ea99d9afb45ba309f58b2312c44a67b491e61db08c79b17b8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:43:23.181251Z","signature_b64":"OPFhHeUSP/f5p3jmRUHJrnThtOzTfibSyctCPU6lXfv5ylFqbe8Ty4Q1s2gCpOPTuD0bsB2pNDmZgsuI1yhjDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"98377cecc293de9b533030c68fe4567ddb4a659d730313c9753f35447256a421","last_reissued_at":"2026-05-17T23:43:23.180675Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:43:23.180675Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distributionally Robust Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Elena Smirnova, Elvis Dohmatob, J\\'er\\'emie Mary","submitted_at":"2019-02-23T00:13:42Z","abstract_excerpt":"Real-world applications require RL algorithms to act safely. During learning process, it is likely that the agent executes sub-optimal actions that may lead to unsafe/poor states of the system. Exploration is particularly brittle in high-dimensional state/action space due to increased number of low-performing actions. In this work, we consider risk-averse exploration in approximate RL setting. To ensure safety during learning, we propose the distributionally robust policy iteration scheme that provides lower bound guarantee on state-values. Our approach induces a dynamic level of risk to preve"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1902.08708","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":""},"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":"1902.08708","created_at":"2026-05-17T23:43:23.180755+00:00"},{"alias_kind":"arxiv_version","alias_value":"1902.08708v2","created_at":"2026-05-17T23:43:23.180755+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1902.08708","created_at":"2026-05-17T23:43:23.180755+00:00"},{"alias_kind":"pith_short_12","alias_value":"TA3XZ3GCSPPJ","created_at":"2026-05-18T12:33:27.125529+00:00"},{"alias_kind":"pith_short_16","alias_value":"TA3XZ3GCSPPJWUZQ","created_at":"2026-05-18T12:33:27.125529+00:00"},{"alias_kind":"pith_short_8","alias_value":"TA3XZ3GC","created_at":"2026-05-18T12:33:27.125529+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2506.12622","citing_title":"DR-SAC: Distributionally Robust Soft Actor-Critic for Reinforcement Learning under Uncertainty","ref_index":47,"is_internal_anchor":true},{"citing_arxiv_id":"2602.08813","citing_title":"Robust Policy Optimization to Prevent Catastrophic Forgetting","ref_index":49,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX","json":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX.json","graph_json":"https://pith.science/api/pith-number/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/graph.json","events_json":"https://pith.science/api/pith-number/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/events.json","paper":"https://pith.science/paper/TA3XZ3GC"},"agent_actions":{"view_html":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX","download_json":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX.json","view_paper":"https://pith.science/paper/TA3XZ3GC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1902.08708&json=true","fetch_graph":"https://pith.science/api/pith-number/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/graph.json","fetch_events":"https://pith.science/api/pith-number/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/action/storage_attestation","attest_author":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/action/author_attestation","sign_citation":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/action/citation_signature","submit_replication":"https://pith.science/pith/TA3XZ3GCSPPJWUZQGDDI7ZCWPX/action/replication_record"}},"created_at":"2026-05-17T23:43:23.180755+00:00","updated_at":"2026-05-17T23:43:23.180755+00:00"}