{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:WRV45YDIVZSPYOXPVJ63KCP75Y","short_pith_number":"pith:WRV45YDI","schema_version":"1.0","canonical_sha256":"b46bcee068ae64fc3aefaa7db509ffee134ad29e0b35e42bf85273f56537eecf","source":{"kind":"arxiv","id":"2310.02951","version":3},"attestation_state":"computed","paper":{"title":"A Fisher-Rao gradient flow for entropy-regularised Markov decision processes in Polish spaces","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.PR"],"primary_cat":"math.OC","authors_text":"Bekzhan Kerimkulov, David Siska, James-Michael Leahy, Lukasz Szpruch, Yufei Zhang","submitted_at":"2023-10-04T16:41:36Z","abstract_excerpt":"We study the global convergence of a Fisher-Rao policy gradient flow for infinite-horizon entropy-regularised Markov decision processes with Polish state and action space. The flow is a continuous-time analogue of a policy mirror descent method. We establish the global well-posedness of the gradient flow and demonstrate its exponential convergence to the optimal policy. Moreover, we prove the flow is stable with respect to gradient evaluation, offering insights into the performance of a natural policy gradient flow with log-linear policy parameterisation. To overcome challenges stemming from t"},"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":"2310.02951","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2023-10-04T16:41:36Z","cross_cats_sorted":["cs.LG","math.PR"],"title_canon_sha256":"eda043d53d44762f0628f945e5316cdabee10368ac3b8e27164e395ce7283daf","abstract_canon_sha256":"b5a0940da4bd4c1d5ef31c82bb2d91631f5135d0e32c2334c199e79fcffda0d6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:47.910090Z","signature_b64":"6woujGOmSHXLJNm2rTOOsaICJ7gQoz9g9YGTkaU7WAEkTYpQ9Lb63fredp9kVQzrdnHp2KYSzVM1PHpzv3pXBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b46bcee068ae64fc3aefaa7db509ffee134ad29e0b35e42bf85273f56537eecf","last_reissued_at":"2026-07-05T11:16:47.909624Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:47.909624Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Fisher-Rao gradient flow for entropy-regularised Markov decision processes in Polish spaces","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.PR"],"primary_cat":"math.OC","authors_text":"Bekzhan Kerimkulov, David Siska, James-Michael Leahy, Lukasz Szpruch, Yufei Zhang","submitted_at":"2023-10-04T16:41:36Z","abstract_excerpt":"We study the global convergence of a Fisher-Rao policy gradient flow for infinite-horizon entropy-regularised Markov decision processes with Polish state and action space. The flow is a continuous-time analogue of a policy mirror descent method. We establish the global well-posedness of the gradient flow and demonstrate its exponential convergence to the optimal policy. Moreover, we prove the flow is stable with respect to gradient evaluation, offering insights into the performance of a natural policy gradient flow with log-linear policy parameterisation. To overcome challenges stemming from t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.02951","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/2310.02951/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":"2310.02951","created_at":"2026-07-05T11:16:47.909679+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.02951v3","created_at":"2026-07-05T11:16:47.909679+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.02951","created_at":"2026-07-05T11:16:47.909679+00:00"},{"alias_kind":"pith_short_12","alias_value":"WRV45YDIVZSP","created_at":"2026-07-05T11:16:47.909679+00:00"},{"alias_kind":"pith_short_16","alias_value":"WRV45YDIVZSPYOXP","created_at":"2026-07-05T11:16:47.909679+00:00"},{"alias_kind":"pith_short_8","alias_value":"WRV45YDI","created_at":"2026-07-05T11:16:47.909679+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04335","citing_title":"Policy Gradient for Continuous-Time Robust Markov Decision Processes","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2402.08106","citing_title":"Mirror Descent-Ascent for mean-field min-max problems","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y","json":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y.json","graph_json":"https://pith.science/api/pith-number/WRV45YDIVZSPYOXPVJ63KCP75Y/graph.json","events_json":"https://pith.science/api/pith-number/WRV45YDIVZSPYOXPVJ63KCP75Y/events.json","paper":"https://pith.science/paper/WRV45YDI"},"agent_actions":{"view_html":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y","download_json":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y.json","view_paper":"https://pith.science/paper/WRV45YDI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.02951&json=true","fetch_graph":"https://pith.science/api/pith-number/WRV45YDIVZSPYOXPVJ63KCP75Y/graph.json","fetch_events":"https://pith.science/api/pith-number/WRV45YDIVZSPYOXPVJ63KCP75Y/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y/action/storage_attestation","attest_author":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y/action/author_attestation","sign_citation":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y/action/citation_signature","submit_replication":"https://pith.science/pith/WRV45YDIVZSPYOXPVJ63KCP75Y/action/replication_record"}},"created_at":"2026-07-05T11:16:47.909679+00:00","updated_at":"2026-07-05T11:16:47.909679+00:00"}