{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LMBEHSRAF75QTK7YJJOVDBEAKV","short_pith_number":"pith:LMBEHSRA","schema_version":"1.0","canonical_sha256":"5b0243ca202ffb09abf84a5d518480555a138adbcd8bcb7b11ea770e7407081d","source":{"kind":"arxiv","id":"2602.06014","version":2},"attestation_state":"computed","paper":{"title":"Optimism Stabilizes Thompson Sampling for Adaptive Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","math.ST","stat.ML","stat.TH"],"primary_cat":"cs.LG","authors_text":"Han Zhong, Shunxing Yan","submitted_at":"2026-02-05T18:52:54Z","abstract_excerpt":"Thompson sampling (TS) is widely used for stochastic multi-armed bandits, yet its inferential properties under adaptive data collection are subtle. Classical asymptotic theory for sample means can fail because arm-specific sample sizes are random and coupled with the rewards through the action-selection rule. We study adaptive inference for Thompson sampling with Gaussian randomized indices in $K$-armed stochastic bandits with independent sub-Gaussian reward noises, and identify \\emph{optimism} as a key mechanism for restoring \\emph{stability}, meaning that each arm's pull count concentrates 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":"2602.06014","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-05T18:52:54Z","cross_cats_sorted":["cs.AI","math.OC","math.ST","stat.ML","stat.TH"],"title_canon_sha256":"5805c8e5603ec5c4694e54304413f8f01572d465c9f88d4d837f27795053241b","abstract_canon_sha256":"71b48e32ba4644e4e5ad784251a475ff7f1c1760e5f1efec9ee1007cf0ac183f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:09:54.510394Z","signature_b64":"kR3E2pYwsPvv7WVvydnisDpwqf7txAkiMDAjcGjQ1pwNvdAmFfX6kfIGbiMMEsO/aTh1+GvT05ifwvQ5ySiiBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b0243ca202ffb09abf84a5d518480555a138adbcd8bcb7b11ea770e7407081d","last_reissued_at":"2026-06-19T16:09:54.509907Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:09:54.509907Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Optimism Stabilizes Thompson Sampling for Adaptive Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","math.ST","stat.ML","stat.TH"],"primary_cat":"cs.LG","authors_text":"Han Zhong, Shunxing Yan","submitted_at":"2026-02-05T18:52:54Z","abstract_excerpt":"Thompson sampling (TS) is widely used for stochastic multi-armed bandits, yet its inferential properties under adaptive data collection are subtle. Classical asymptotic theory for sample means can fail because arm-specific sample sizes are random and coupled with the rewards through the action-selection rule. We study adaptive inference for Thompson sampling with Gaussian randomized indices in $K$-armed stochastic bandits with independent sub-Gaussian reward noises, and identify \\emph{optimism} as a key mechanism for restoring \\emph{stability}, meaning that each arm's pull count concentrates a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.06014","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/2602.06014/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":"2602.06014","created_at":"2026-06-19T16:09:54.509965+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.06014v2","created_at":"2026-06-19T16:09:54.509965+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.06014","created_at":"2026-06-19T16:09:54.509965+00:00"},{"alias_kind":"pith_short_12","alias_value":"LMBEHSRAF75Q","created_at":"2026-06-19T16:09:54.509965+00:00"},{"alias_kind":"pith_short_16","alias_value":"LMBEHSRAF75QTK7Y","created_at":"2026-06-19T16:09:54.509965+00:00"},{"alias_kind":"pith_short_8","alias_value":"LMBEHSRA","created_at":"2026-06-19T16:09:54.509965+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.00913","citing_title":"Bandit Simulation for Average Reward Inference","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV","json":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV.json","graph_json":"https://pith.science/api/pith-number/LMBEHSRAF75QTK7YJJOVDBEAKV/graph.json","events_json":"https://pith.science/api/pith-number/LMBEHSRAF75QTK7YJJOVDBEAKV/events.json","paper":"https://pith.science/paper/LMBEHSRA"},"agent_actions":{"view_html":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV","download_json":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV.json","view_paper":"https://pith.science/paper/LMBEHSRA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.06014&json=true","fetch_graph":"https://pith.science/api/pith-number/LMBEHSRAF75QTK7YJJOVDBEAKV/graph.json","fetch_events":"https://pith.science/api/pith-number/LMBEHSRAF75QTK7YJJOVDBEAKV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV/action/storage_attestation","attest_author":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV/action/author_attestation","sign_citation":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV/action/citation_signature","submit_replication":"https://pith.science/pith/LMBEHSRAF75QTK7YJJOVDBEAKV/action/replication_record"}},"created_at":"2026-06-19T16:09:54.509965+00:00","updated_at":"2026-06-19T16:09:54.509965+00:00"}