{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:4IXMM3LGXQT7QHWT23LNGXBWVE","short_pith_number":"pith:4IXMM3LG","schema_version":"1.0","canonical_sha256":"e22ec66d66bc27f81ed3d6d6d35c36a93144991974ded528f1b21767ff10ca90","source":{"kind":"arxiv","id":"2505.12462","version":3},"attestation_state":"computed","paper":{"title":"Model-Free Robust Average-Reward Reinforcement Learning with Sample Complexity Analysis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"George Atia, Yue Wang, Zachary Roch","submitted_at":"2025-05-18T15:34:45Z","abstract_excerpt":"Robust reinforcement learning (RL) under the average-reward criterion is essential for long-term decision-making, particularly when the environment may differ from its training dynamics. However, most existing studies focus on model-based settings and provide only asymptotic guarantees, hindering their principled understanding and practical deployment, especially in data-limited scenarios. We aim to close this gap by proposing a model-free algorithm, \\textbf{Robust Halpern Iteration (RHI)}. We first design our algorithm based on a black-box sampling oracle, which can estimate the worst-case pe"},"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":"2505.12462","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-18T15:34:45Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"972691b02cc1d33deaa10834e9cd5ed1081b3643aea06a25763b63a379d0866a","abstract_canon_sha256":"1acc72e398a0f3919f79b1408f0f2823f70937355d6ef6c7cc7a86eb72870113"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T02:13:15.305742Z","signature_b64":"o/pwYFz4awc5rzt+S8naj83QtDt83odY8cxyGLW7yfB+cfyAaaiD9tY5mhYYQFhab1cUmeNo8oDNEfH7Upj1Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e22ec66d66bc27f81ed3d6d6d35c36a93144991974ded528f1b21767ff10ca90","last_reissued_at":"2026-06-23T02:13:15.305321Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T02:13:15.305321Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Model-Free Robust Average-Reward Reinforcement Learning with Sample Complexity Analysis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"George Atia, Yue Wang, Zachary Roch","submitted_at":"2025-05-18T15:34:45Z","abstract_excerpt":"Robust reinforcement learning (RL) under the average-reward criterion is essential for long-term decision-making, particularly when the environment may differ from its training dynamics. However, most existing studies focus on model-based settings and provide only asymptotic guarantees, hindering their principled understanding and practical deployment, especially in data-limited scenarios. We aim to close this gap by proposing a model-free algorithm, \\textbf{Robust Halpern Iteration (RHI)}. We first design our algorithm based on a black-box sampling oracle, which can estimate the worst-case pe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12462","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/2505.12462/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":"2505.12462","created_at":"2026-06-23T02:13:15.305386+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.12462v3","created_at":"2026-06-23T02:13:15.305386+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12462","created_at":"2026-06-23T02:13:15.305386+00:00"},{"alias_kind":"pith_short_12","alias_value":"4IXMM3LGXQT7","created_at":"2026-06-23T02:13:15.305386+00:00"},{"alias_kind":"pith_short_16","alias_value":"4IXMM3LGXQT7QHWT","created_at":"2026-06-23T02:13:15.305386+00:00"},{"alias_kind":"pith_short_8","alias_value":"4IXMM3LG","created_at":"2026-06-23T02:13:15.305386+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.07040","citing_title":"Efficient Q-Learning and Actor-Critic Methods for Robust Average-Reward Reinforcement Learning","ref_index":16,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE","json":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE.json","graph_json":"https://pith.science/api/pith-number/4IXMM3LGXQT7QHWT23LNGXBWVE/graph.json","events_json":"https://pith.science/api/pith-number/4IXMM3LGXQT7QHWT23LNGXBWVE/events.json","paper":"https://pith.science/paper/4IXMM3LG"},"agent_actions":{"view_html":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE","download_json":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE.json","view_paper":"https://pith.science/paper/4IXMM3LG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.12462&json=true","fetch_graph":"https://pith.science/api/pith-number/4IXMM3LGXQT7QHWT23LNGXBWVE/graph.json","fetch_events":"https://pith.science/api/pith-number/4IXMM3LGXQT7QHWT23LNGXBWVE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE/action/storage_attestation","attest_author":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE/action/author_attestation","sign_citation":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE/action/citation_signature","submit_replication":"https://pith.science/pith/4IXMM3LGXQT7QHWT23LNGXBWVE/action/replication_record"}},"created_at":"2026-06-23T02:13:15.305386+00:00","updated_at":"2026-06-23T02:13:15.305386+00:00"}