{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:Y247JZGQFIZ6X5RJX4ITGVRZOF","short_pith_number":"pith:Y247JZGQ","schema_version":"1.0","canonical_sha256":"c6b9f4e4d02a33ebf629bf11335639715699eb5e7c7add4556b79a899dc3a56f","source":{"kind":"arxiv","id":"2306.05991","version":1},"attestation_state":"computed","paper":{"title":"Approximate information state based convergence analysis of recurrent Q-learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aditya Mahajan, Amit Sinha, Erfan Seyedsalehi, Nima Akbarzadeh","submitted_at":"2023-06-09T15:59:39Z","abstract_excerpt":"In spite of the large literature on reinforcement learning (RL) algorithms for partially observable Markov decision processes (POMDPs), a complete theoretical understanding is still lacking. In a partially observable setting, the history of data available to the agent increases over time so most practical algorithms either truncate the history to a finite window or compress it using a recurrent neural network leading to an agent state that is non-Markovian. In this paper, it is shown that in spite of the lack of the Markov property, recurrent Q-learning (RQL) converges in the tabular setting. "},"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":"2306.05991","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-06-09T15:59:39Z","cross_cats_sorted":[],"title_canon_sha256":"9ca720906c39ce4fcdc3e8b9073ede43cd01ded6b22d5ba4409cb6772651c8d5","abstract_canon_sha256":"8f9f1dc34a2dd97ecfe6c0e061279cd139b9f3158ba7ccbdae35af059ef0cc98"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:19:12.599940Z","signature_b64":"7itl/hRQpGGOAeEPlvaqSgqANyhbDVjSd+biZgqd6ais+NyaGJT9hCqfW8dcQ53gmvICdl2ywHyviF6E+Pq4DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c6b9f4e4d02a33ebf629bf11335639715699eb5e7c7add4556b79a899dc3a56f","last_reissued_at":"2026-07-05T06:19:12.599531Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:19:12.599531Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Approximate information state based convergence analysis of recurrent Q-learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aditya Mahajan, Amit Sinha, Erfan Seyedsalehi, Nima Akbarzadeh","submitted_at":"2023-06-09T15:59:39Z","abstract_excerpt":"In spite of the large literature on reinforcement learning (RL) algorithms for partially observable Markov decision processes (POMDPs), a complete theoretical understanding is still lacking. In a partially observable setting, the history of data available to the agent increases over time so most practical algorithms either truncate the history to a finite window or compress it using a recurrent neural network leading to an agent state that is non-Markovian. In this paper, it is shown that in spite of the lack of the Markov property, recurrent Q-learning (RQL) converges in the tabular setting. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.05991","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/2306.05991/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":"2306.05991","created_at":"2026-07-05T06:19:12.599591+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.05991v1","created_at":"2026-07-05T06:19:12.599591+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.05991","created_at":"2026-07-05T06:19:12.599591+00:00"},{"alias_kind":"pith_short_12","alias_value":"Y247JZGQFIZ6","created_at":"2026-07-05T06:19:12.599591+00:00"},{"alias_kind":"pith_short_16","alias_value":"Y247JZGQFIZ6X5RJ","created_at":"2026-07-05T06:19:12.599591+00:00"},{"alias_kind":"pith_short_8","alias_value":"Y247JZGQ","created_at":"2026-07-05T06:19:12.599591+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.10816","citing_title":"Policy Gradient Methods for Non-Markovian Reinforcement Learning","ref_index":45,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF","json":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF.json","graph_json":"https://pith.science/api/pith-number/Y247JZGQFIZ6X5RJX4ITGVRZOF/graph.json","events_json":"https://pith.science/api/pith-number/Y247JZGQFIZ6X5RJX4ITGVRZOF/events.json","paper":"https://pith.science/paper/Y247JZGQ"},"agent_actions":{"view_html":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF","download_json":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF.json","view_paper":"https://pith.science/paper/Y247JZGQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.05991&json=true","fetch_graph":"https://pith.science/api/pith-number/Y247JZGQFIZ6X5RJX4ITGVRZOF/graph.json","fetch_events":"https://pith.science/api/pith-number/Y247JZGQFIZ6X5RJX4ITGVRZOF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF/action/storage_attestation","attest_author":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF/action/author_attestation","sign_citation":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF/action/citation_signature","submit_replication":"https://pith.science/pith/Y247JZGQFIZ6X5RJX4ITGVRZOF/action/replication_record"}},"created_at":"2026-07-05T06:19:12.599591+00:00","updated_at":"2026-07-05T06:19:12.599591+00:00"}