{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2M4SWOJVPQHLPDY5CVOZ55KQTM","short_pith_number":"pith:2M4SWOJV","schema_version":"1.0","canonical_sha256":"d3392b39357c0eb78f1d155d9ef5509b0197c0bc7aa191da8cd622dfed4ca867","source":{"kind":"arxiv","id":"2505.23244","version":2},"attestation_state":"computed","paper":{"title":"Equivalence of stochastic and deterministic policy gradients","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Emo Todorov","submitted_at":"2025-05-29T08:51:34Z","abstract_excerpt":"Policy gradients in continuous control have been derived for both stochastic and deterministic policies. Here we study the relationship between the two. In a widely-used family of MDPs involving Gaussian control noise and quadratic control costs, we show that the stochastic and deterministic policy gradients, natural gradients, and state value functions are identical; while the state-control value functions are different. We then develop a general procedure for constructing an MDP with deterministic policy that is equivalent to a given MDP with stochastic policy. The controls of this new MDP 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":"2505.23244","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-29T08:51:34Z","cross_cats_sorted":[],"title_canon_sha256":"737f307dba97b1a9c86041f44b8de6d1b645d8adc06f5eb5b0c1324d561159ef","abstract_canon_sha256":"dc9de14db70edea9ca9ccbcc6fa181444a0303c17049864bd1cb0767b8f71976"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:43.333192Z","signature_b64":"D/qFwFWfc+R0tzXjrs1wIlUKoXBUpej3iXJ999pmrdxct6de8ZAaJ0QlzfITKiLrB3pkUw/s5DWqXLahAfdrCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d3392b39357c0eb78f1d155d9ef5509b0197c0bc7aa191da8cd622dfed4ca867","last_reissued_at":"2026-07-05T11:18:43.332711Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:43.332711Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Equivalence of stochastic and deterministic policy gradients","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Emo Todorov","submitted_at":"2025-05-29T08:51:34Z","abstract_excerpt":"Policy gradients in continuous control have been derived for both stochastic and deterministic policies. Here we study the relationship between the two. In a widely-used family of MDPs involving Gaussian control noise and quadratic control costs, we show that the stochastic and deterministic policy gradients, natural gradients, and state value functions are identical; while the state-control value functions are different. We then develop a general procedure for constructing an MDP with deterministic policy that is equivalent to a given MDP with stochastic policy. The controls of this new MDP a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.23244","kind":"arxiv","version":2},"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/2505.23244/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":"2505.23244","created_at":"2026-07-05T11:18:43.332766+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.23244v2","created_at":"2026-07-05T11:18:43.332766+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.23244","created_at":"2026-07-05T11:18:43.332766+00:00"},{"alias_kind":"pith_short_12","alias_value":"2M4SWOJVPQHL","created_at":"2026-07-05T11:18:43.332766+00:00"},{"alias_kind":"pith_short_16","alias_value":"2M4SWOJVPQHLPDY5","created_at":"2026-07-05T11:18:43.332766+00:00"},{"alias_kind":"pith_short_8","alias_value":"2M4SWOJV","created_at":"2026-07-05T11:18:43.332766+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/2M4SWOJVPQHLPDY5CVOZ55KQTM","json":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM.json","graph_json":"https://pith.science/api/pith-number/2M4SWOJVPQHLPDY5CVOZ55KQTM/graph.json","events_json":"https://pith.science/api/pith-number/2M4SWOJVPQHLPDY5CVOZ55KQTM/events.json","paper":"https://pith.science/paper/2M4SWOJV"},"agent_actions":{"view_html":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM","download_json":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM.json","view_paper":"https://pith.science/paper/2M4SWOJV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.23244&json=true","fetch_graph":"https://pith.science/api/pith-number/2M4SWOJVPQHLPDY5CVOZ55KQTM/graph.json","fetch_events":"https://pith.science/api/pith-number/2M4SWOJVPQHLPDY5CVOZ55KQTM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM/action/storage_attestation","attest_author":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM/action/author_attestation","sign_citation":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM/action/citation_signature","submit_replication":"https://pith.science/pith/2M4SWOJVPQHLPDY5CVOZ55KQTM/action/replication_record"}},"created_at":"2026-07-05T11:18:43.332766+00:00","updated_at":"2026-07-05T11:18:43.332766+00:00"}