{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:4FNQIT5LH7CMMEPIA7IRCOROLN","short_pith_number":"pith:4FNQIT5L","schema_version":"1.0","canonical_sha256":"e15b044fab3fc4c611e807d1113a2e5b629594de803abed4978a2094be56d14d","source":{"kind":"arxiv","id":"2305.18246","version":2},"attestation_state":"computed","paper":{"title":"Provable and Practical: Efficient Exploration in Reinforcement Learning via Langevin Monte Carlo","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anima Anandkumar, A. Rupam Mahmood, Doina Precup, Haque Ishfaq, Kamyar Azizzadenesheli, Pan Xu, Qingfeng Lan","submitted_at":"2023-05-29T17:11:28Z","abstract_excerpt":"We present a scalable and effective exploration strategy based on Thompson sampling for reinforcement learning (RL). One of the key shortcomings of existing Thompson sampling algorithms is the need to perform a Gaussian approximation of the posterior distribution, which is not a good surrogate in most practical settings. We instead directly sample the Q function from its posterior distribution, by using Langevin Monte Carlo, an efficient type of Markov Chain Monte Carlo (MCMC) method. Our method only needs to perform noisy gradient descent updates to learn the exact posterior distribution of 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":"2305.18246","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-05-29T17:11:28Z","cross_cats_sorted":[],"title_canon_sha256":"cf8219e2c1d5943e24a3a6d3fe1e993ed3d1a2636690141d5c81799167d16501","abstract_canon_sha256":"0bb3ebb52526d08aa73df1337dff9e242252fda4e52123f1639bbdc9418d7bc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:56:48.830079Z","signature_b64":"fDrfQ9i8fiYYS81jUBjMq1YhYOB20u3BqnSZyPBPluFOAsED34qDAjocArfNP4+eUouDiej3mIl8NVzYlcRmAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e15b044fab3fc4c611e807d1113a2e5b629594de803abed4978a2094be56d14d","last_reissued_at":"2026-07-05T07:56:48.829594Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:56:48.829594Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Provable and Practical: Efficient Exploration in Reinforcement Learning via Langevin Monte Carlo","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anima Anandkumar, A. Rupam Mahmood, Doina Precup, Haque Ishfaq, Kamyar Azizzadenesheli, Pan Xu, Qingfeng Lan","submitted_at":"2023-05-29T17:11:28Z","abstract_excerpt":"We present a scalable and effective exploration strategy based on Thompson sampling for reinforcement learning (RL). One of the key shortcomings of existing Thompson sampling algorithms is the need to perform a Gaussian approximation of the posterior distribution, which is not a good surrogate in most practical settings. We instead directly sample the Q function from its posterior distribution, by using Langevin Monte Carlo, an efficient type of Markov Chain Monte Carlo (MCMC) method. Our method only needs to perform noisy gradient descent updates to learn the exact posterior distribution of t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.18246","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/2305.18246/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":"2305.18246","created_at":"2026-07-05T07:56:48.829652+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.18246v2","created_at":"2026-07-05T07:56:48.829652+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.18246","created_at":"2026-07-05T07:56:48.829652+00:00"},{"alias_kind":"pith_short_12","alias_value":"4FNQIT5LH7CM","created_at":"2026-07-05T07:56:48.829652+00:00"},{"alias_kind":"pith_short_16","alias_value":"4FNQIT5LH7CMMEPI","created_at":"2026-07-05T07:56:48.829652+00:00"},{"alias_kind":"pith_short_8","alias_value":"4FNQIT5L","created_at":"2026-07-05T07:56:48.829652+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2602.06239","citing_title":"Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2602.06239","citing_title":"Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN","json":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN.json","graph_json":"https://pith.science/api/pith-number/4FNQIT5LH7CMMEPIA7IRCOROLN/graph.json","events_json":"https://pith.science/api/pith-number/4FNQIT5LH7CMMEPIA7IRCOROLN/events.json","paper":"https://pith.science/paper/4FNQIT5L"},"agent_actions":{"view_html":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN","download_json":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN.json","view_paper":"https://pith.science/paper/4FNQIT5L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.18246&json=true","fetch_graph":"https://pith.science/api/pith-number/4FNQIT5LH7CMMEPIA7IRCOROLN/graph.json","fetch_events":"https://pith.science/api/pith-number/4FNQIT5LH7CMMEPIA7IRCOROLN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN/action/storage_attestation","attest_author":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN/action/author_attestation","sign_citation":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN/action/citation_signature","submit_replication":"https://pith.science/pith/4FNQIT5LH7CMMEPIA7IRCOROLN/action/replication_record"}},"created_at":"2026-07-05T07:56:48.829652+00:00","updated_at":"2026-07-05T07:56:48.829652+00:00"}