{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:C33AFYNRS7X2FNGROROZNUUKNS","short_pith_number":"pith:C33AFYNR","schema_version":"1.0","canonical_sha256":"16f602e1b197efa2b4d1745d96d28a6ca0415da9aaf96eba26125e2ab68b0592","source":{"kind":"arxiv","id":"2605.24939","version":1},"attestation_state":"computed","paper":{"title":"Global linear convergence of entropy-regularized softmax policy gradient beyond tabular MDPs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"David \\v{S}i\\v{s}ka, Lukasz Szpruch, Ziyue Chen","submitted_at":"2026-05-24T08:38:29Z","abstract_excerpt":"We study the global convergence of policy gradient for infinite-horizon entropy-regularized Markov decision processes (MDPs) with continuous state and action spaces. We consider log-linear softmax policies with linear function approximation, which extend the tabular softmax parameterization while retaining a tractable policy class. Under $Q^\\pi_\\tau$-realizability for the regularized state-action value function, we first establish a non-uniform Polyak--{\\L}ojasiewicz (P\\L) inequality. The non-uniformity arises through degeneracy of constants associated with the policy geometry, namely the Fish"},"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":"2605.24939","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-24T08:38:29Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"5f45e4b53d8c4eee0aff1ad5cc978869b45dae2130924c1f9248b28c3c5b9603","abstract_canon_sha256":"871483254df9c451c06829892c662c34cb122c8bfa2f53b80e13e5613ffcee0a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:04:06.740971Z","signature_b64":"YYACD2EW1ioWzcctETLwaUHLJK9nYkJH8YhekkUusHSjuRv2AdhCSaXt1L+7pm0uDp5JO6C1GYoYA1UWcJDoCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"16f602e1b197efa2b4d1745d96d28a6ca0415da9aaf96eba26125e2ab68b0592","last_reissued_at":"2026-05-26T01:04:06.740223Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:04:06.740223Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Global linear convergence of entropy-regularized softmax policy gradient beyond tabular MDPs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"David \\v{S}i\\v{s}ka, Lukasz Szpruch, Ziyue Chen","submitted_at":"2026-05-24T08:38:29Z","abstract_excerpt":"We study the global convergence of policy gradient for infinite-horizon entropy-regularized Markov decision processes (MDPs) with continuous state and action spaces. We consider log-linear softmax policies with linear function approximation, which extend the tabular softmax parameterization while retaining a tractable policy class. Under $Q^\\pi_\\tau$-realizability for the regularized state-action value function, we first establish a non-uniform Polyak--{\\L}ojasiewicz (P\\L) inequality. The non-uniformity arises through degeneracy of constants associated with the policy geometry, namely the Fish"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.24939","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/2605.24939/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":"2605.24939","created_at":"2026-05-26T01:04:06.740347+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.24939v1","created_at":"2026-05-26T01:04:06.740347+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.24939","created_at":"2026-05-26T01:04:06.740347+00:00"},{"alias_kind":"pith_short_12","alias_value":"C33AFYNRS7X2","created_at":"2026-05-26T01:04:06.740347+00:00"},{"alias_kind":"pith_short_16","alias_value":"C33AFYNRS7X2FNGR","created_at":"2026-05-26T01:04:06.740347+00:00"},{"alias_kind":"pith_short_8","alias_value":"C33AFYNR","created_at":"2026-05-26T01:04:06.740347+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS","json":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS.json","graph_json":"https://pith.science/api/pith-number/C33AFYNRS7X2FNGROROZNUUKNS/graph.json","events_json":"https://pith.science/api/pith-number/C33AFYNRS7X2FNGROROZNUUKNS/events.json","paper":"https://pith.science/paper/C33AFYNR"},"agent_actions":{"view_html":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS","download_json":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS.json","view_paper":"https://pith.science/paper/C33AFYNR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.24939&json=true","fetch_graph":"https://pith.science/api/pith-number/C33AFYNRS7X2FNGROROZNUUKNS/graph.json","fetch_events":"https://pith.science/api/pith-number/C33AFYNRS7X2FNGROROZNUUKNS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS/action/storage_attestation","attest_author":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS/action/author_attestation","sign_citation":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS/action/citation_signature","submit_replication":"https://pith.science/pith/C33AFYNRS7X2FNGROROZNUUKNS/action/replication_record"}},"created_at":"2026-05-26T01:04:06.740347+00:00","updated_at":"2026-05-26T01:04:06.740347+00:00"}