{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YJ6I6UYLZCH2MWUNPSAXZZDOZ5","short_pith_number":"pith:YJ6I6UYL","schema_version":"1.0","canonical_sha256":"c27c8f530bc88fa65a8d7c817ce46ecf45b6c2ab6490c106aa4efe7e603af8b9","source":{"kind":"arxiv","id":"2506.00286","version":3},"attestation_state":"computed","paper":{"title":"Recursive Entropic Risk Optimization in Discounted MDPs: Sample Complexity Bounds with a Generative Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Mohammad Sadegh Talebi, Oliver Mortensen","submitted_at":"2025-05-30T22:27:57Z","abstract_excerpt":"We study risk-sensitive reinforcement learning in finite discounted MDPs with recursive entropic risk measures (ERM), where the risk parameter $\\beta \\neq 0$ controls the agent's risk attitude: $\\beta>0$ for risk-averse and $\\beta<0$ for risk-seeking behavior. A generative model of the MDP is assumed to be available. Our focus is on the sample complexities of learning the optimal state-action value function (value learning) and an optimal policy (policy learning) under recursive ERM. We introduce a model-based algorithm, called Model-Based ERM $Q$-Value Iteration (MB-RS-QVI), and derive PAC-ty"},"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":"2506.00286","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T22:27:57Z","cross_cats_sorted":["cs.AI","math.OC","stat.ML"],"title_canon_sha256":"8c78ee9bba21ddf364436b41d110d1dfa6482010288e88b2ea52dd237f336841","abstract_canon_sha256":"4991c51d5b974b4dc3d8483a86051af27416d7c4fad226c5c3316f99589ca7aa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:04:54.516232Z","signature_b64":"CrQaiiAN8kc6q9X0YWx/TU+TcimQC7SA4ayhUAxaU89cq9TKbPm3auZPHgIw4Cq4D0KRmQkDd+NIfJQcCTvCCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c27c8f530bc88fa65a8d7c817ce46ecf45b6c2ab6490c106aa4efe7e603af8b9","last_reissued_at":"2026-05-20T01:04:54.515290Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:04:54.515290Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Recursive Entropic Risk Optimization in Discounted MDPs: Sample Complexity Bounds with a Generative Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Mohammad Sadegh Talebi, Oliver Mortensen","submitted_at":"2025-05-30T22:27:57Z","abstract_excerpt":"We study risk-sensitive reinforcement learning in finite discounted MDPs with recursive entropic risk measures (ERM), where the risk parameter $\\beta \\neq 0$ controls the agent's risk attitude: $\\beta>0$ for risk-averse and $\\beta<0$ for risk-seeking behavior. A generative model of the MDP is assumed to be available. Our focus is on the sample complexities of learning the optimal state-action value function (value learning) and an optimal policy (policy learning) under recursive ERM. We introduce a model-based algorithm, called Model-Based ERM $Q$-Value Iteration (MB-RS-QVI), and derive PAC-ty"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.00286","kind":"arxiv","version":3},"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/2506.00286/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":"2506.00286","created_at":"2026-05-20T01:04:54.515421+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.00286v3","created_at":"2026-05-20T01:04:54.515421+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.00286","created_at":"2026-05-20T01:04:54.515421+00:00"},{"alias_kind":"pith_short_12","alias_value":"YJ6I6UYLZCH2","created_at":"2026-05-20T01:04:54.515421+00:00"},{"alias_kind":"pith_short_16","alias_value":"YJ6I6UYLZCH2MWUN","created_at":"2026-05-20T01:04:54.515421+00:00"},{"alias_kind":"pith_short_8","alias_value":"YJ6I6UYL","created_at":"2026-05-20T01:04:54.515421+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.21763","citing_title":"On the Sample Complexity of Discounted Reinforcement Learning with Optimized Certainty Equivalents","ref_index":52,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13717","citing_title":"Tight Sample Complexity Bounds for Entropic Best Policy Identification","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5","json":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5.json","graph_json":"https://pith.science/api/pith-number/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/graph.json","events_json":"https://pith.science/api/pith-number/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/events.json","paper":"https://pith.science/paper/YJ6I6UYL"},"agent_actions":{"view_html":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5","download_json":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5.json","view_paper":"https://pith.science/paper/YJ6I6UYL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.00286&json=true","fetch_graph":"https://pith.science/api/pith-number/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/graph.json","fetch_events":"https://pith.science/api/pith-number/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/action/storage_attestation","attest_author":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/action/author_attestation","sign_citation":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/action/citation_signature","submit_replication":"https://pith.science/pith/YJ6I6UYLZCH2MWUNPSAXZZDOZ5/action/replication_record"}},"created_at":"2026-05-20T01:04:54.515421+00:00","updated_at":"2026-05-20T01:04:54.515421+00:00"}