{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:5T4KTDRHKKVA37U3DWSYSASRQL","short_pith_number":"pith:5T4KTDRH","schema_version":"1.0","canonical_sha256":"ecf8a98e2752aa0dfe9b1da589025182ea76e52ee4c48175e2936955aa25474d","source":{"kind":"arxiv","id":"1712.07185","version":1},"attestation_state":"computed","paper":{"title":"On Wasserstein Reinforcement Learning and the Fokker-Planck equation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Brendan Maginnis, Pierre H. Richemond","submitted_at":"2017-12-19T20:05:06Z","abstract_excerpt":"Policy gradients methods often achieve better performance when the change in policy is limited to a small Kullback-Leibler divergence. We derive policy gradients where the change in policy is limited to a small Wasserstein distance (or trust region). This is done in the discrete and continuous multi-armed bandit settings with entropy regularisation. We show that in the small steps limit with respect to the Wasserstein distance $W_2$, policy dynamics are governed by the Fokker-Planck (heat) equation, following the Jordan-Kinderlehrer-Otto result. This means that policies undergo diffusion and a"},"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":"1712.07185","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-19T20:05:06Z","cross_cats_sorted":[],"title_canon_sha256":"d090aedc584432b4c717b2149e6dc5a445d19cf29c7a1c380fa969f0fe2e517d","abstract_canon_sha256":"87d163e9fa6188eceed4e5fec662a37a5c96a3ecc936fe7c001d6a181265a814"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:27:35.310575Z","signature_b64":"ZpZ0d+GHQ7q6USRNrIT5KWlkT9c8nn+Day1h2kXF6baxm5l12HsxzrlV2yYsx9GC5irwibhzEF6w226Cy+c5Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ecf8a98e2752aa0dfe9b1da589025182ea76e52ee4c48175e2936955aa25474d","last_reissued_at":"2026-05-18T00:27:35.309945Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:27:35.309945Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Wasserstein Reinforcement Learning and the Fokker-Planck equation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Brendan Maginnis, Pierre H. Richemond","submitted_at":"2017-12-19T20:05:06Z","abstract_excerpt":"Policy gradients methods often achieve better performance when the change in policy is limited to a small Kullback-Leibler divergence. We derive policy gradients where the change in policy is limited to a small Wasserstein distance (or trust region). This is done in the discrete and continuous multi-armed bandit settings with entropy regularisation. We show that in the small steps limit with respect to the Wasserstein distance $W_2$, policy dynamics are governed by the Fokker-Planck (heat) equation, following the Jordan-Kinderlehrer-Otto result. This means that policies undergo diffusion and a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1712.07185","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":""},"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":"1712.07185","created_at":"2026-05-18T00:27:35.310040+00:00"},{"alias_kind":"arxiv_version","alias_value":"1712.07185v1","created_at":"2026-05-18T00:27:35.310040+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1712.07185","created_at":"2026-05-18T00:27:35.310040+00:00"},{"alias_kind":"pith_short_12","alias_value":"5T4KTDRHKKVA","created_at":"2026-05-18T12:31:00.734936+00:00"},{"alias_kind":"pith_short_16","alias_value":"5T4KTDRHKKVA37U3","created_at":"2026-05-18T12:31:00.734936+00:00"},{"alias_kind":"pith_short_8","alias_value":"5T4KTDRH","created_at":"2026-05-18T12:31:00.734936+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.23517","citing_title":"Inexact JKO and proximal-gradient algorithms in the Wasserstein space","ref_index":61,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL","json":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL.json","graph_json":"https://pith.science/api/pith-number/5T4KTDRHKKVA37U3DWSYSASRQL/graph.json","events_json":"https://pith.science/api/pith-number/5T4KTDRHKKVA37U3DWSYSASRQL/events.json","paper":"https://pith.science/paper/5T4KTDRH"},"agent_actions":{"view_html":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL","download_json":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL.json","view_paper":"https://pith.science/paper/5T4KTDRH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1712.07185&json=true","fetch_graph":"https://pith.science/api/pith-number/5T4KTDRHKKVA37U3DWSYSASRQL/graph.json","fetch_events":"https://pith.science/api/pith-number/5T4KTDRHKKVA37U3DWSYSASRQL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL/action/storage_attestation","attest_author":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL/action/author_attestation","sign_citation":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL/action/citation_signature","submit_replication":"https://pith.science/pith/5T4KTDRHKKVA37U3DWSYSASRQL/action/replication_record"}},"created_at":"2026-05-18T00:27:35.310040+00:00","updated_at":"2026-05-18T00:27:35.310040+00:00"}