{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:SQ3LR5H2UHQR2K7NUTOA7IQZYN","short_pith_number":"pith:SQ3LR5H2","schema_version":"1.0","canonical_sha256":"9436b8f4faa1e11d2beda4dc0fa219c34a6d8c38ea64cff5932b6e29ca3c4b51","source":{"kind":"arxiv","id":"1908.04970","version":2},"attestation_state":"computed","paper":{"title":"Thompson Sampling with Approximate Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Justin Domke, My Phan, Yasin Abbasi-Yadkori","submitted_at":"2019-08-14T06:09:15Z","abstract_excerpt":"We study the effects of approximate inference on the performance of Thompson sampling in the $k$-armed bandit problems. Thompson sampling is a successful algorithm for online decision-making but requires posterior inference, which often must be approximated in practice. We show that even small constant inference error (in $\\alpha$-divergence) can lead to poor performance (linear regret) due to under-exploration (for $\\alpha<1$) or over-exploration (for $\\alpha>0$) by the approximation. While for $\\alpha > 0$ this is unavoidable, for $\\alpha \\leq 0$ the regret can be improved by adding a small "},"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":"1908.04970","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-08-14T06:09:15Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"163d28d5e1df3bc928e9b44781172cce6c93915f28ddd2075410a2fb52eec29d","abstract_canon_sha256":"8fb422de6957359e26e71e13df1c25f8f1b06ddd32d72b659aaa91bcd99dd75e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:33:35.306230Z","signature_b64":"q8mqGV4zqPryuetMNa5TyDiapY0kUihVaAmME8rOA1YmnfKQzriPlXv9YFS93HJCI18RwhndTFh4fBHEv3jvCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9436b8f4faa1e11d2beda4dc0fa219c34a6d8c38ea64cff5932b6e29ca3c4b51","last_reissued_at":"2026-07-05T00:33:35.305695Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:33:35.305695Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Thompson Sampling with Approximate Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Justin Domke, My Phan, Yasin Abbasi-Yadkori","submitted_at":"2019-08-14T06:09:15Z","abstract_excerpt":"We study the effects of approximate inference on the performance of Thompson sampling in the $k$-armed bandit problems. Thompson sampling is a successful algorithm for online decision-making but requires posterior inference, which often must be approximated in practice. We show that even small constant inference error (in $\\alpha$-divergence) can lead to poor performance (linear regret) due to under-exploration (for $\\alpha<1$) or over-exploration (for $\\alpha>0$) by the approximation. While for $\\alpha > 0$ this is unavoidable, for $\\alpha \\leq 0$ the regret can be improved by adding a small "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.04970","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/1908.04970/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":"1908.04970","created_at":"2026-07-05T00:33:35.305749+00:00"},{"alias_kind":"arxiv_version","alias_value":"1908.04970v2","created_at":"2026-07-05T00:33:35.305749+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.04970","created_at":"2026-07-05T00:33:35.305749+00:00"},{"alias_kind":"pith_short_12","alias_value":"SQ3LR5H2UHQR","created_at":"2026-07-05T00:33:35.305749+00:00"},{"alias_kind":"pith_short_16","alias_value":"SQ3LR5H2UHQR2K7N","created_at":"2026-07-05T00:33:35.305749+00:00"},{"alias_kind":"pith_short_8","alias_value":"SQ3LR5H2","created_at":"2026-07-05T00:33:35.305749+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/SQ3LR5H2UHQR2K7NUTOA7IQZYN","json":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN.json","graph_json":"https://pith.science/api/pith-number/SQ3LR5H2UHQR2K7NUTOA7IQZYN/graph.json","events_json":"https://pith.science/api/pith-number/SQ3LR5H2UHQR2K7NUTOA7IQZYN/events.json","paper":"https://pith.science/paper/SQ3LR5H2"},"agent_actions":{"view_html":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN","download_json":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN.json","view_paper":"https://pith.science/paper/SQ3LR5H2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1908.04970&json=true","fetch_graph":"https://pith.science/api/pith-number/SQ3LR5H2UHQR2K7NUTOA7IQZYN/graph.json","fetch_events":"https://pith.science/api/pith-number/SQ3LR5H2UHQR2K7NUTOA7IQZYN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN/action/storage_attestation","attest_author":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN/action/author_attestation","sign_citation":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN/action/citation_signature","submit_replication":"https://pith.science/pith/SQ3LR5H2UHQR2K7NUTOA7IQZYN/action/replication_record"}},"created_at":"2026-07-05T00:33:35.305749+00:00","updated_at":"2026-07-05T00:33:35.305749+00:00"}