{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:KNGCKTCKB3YDLDSEBQV4BW5ZKO","short_pith_number":"pith:KNGCKTCK","schema_version":"1.0","canonical_sha256":"534c254c4a0ef0358e440c2bc0dbb953a0026cda0a0ea3954c8052815b066a23","source":{"kind":"arxiv","id":"2604.20657","version":2},"attestation_state":"computed","paper":{"title":"Importance Sampling in Expensive Finite-Sum Optimization via Contextual Bandit Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Framing subset selection for stochastic average model methods as a contextual bandit problem lets the Exp4 algorithm use side information to create better sampling distributions than uniform randomization.","cross_cats":[],"primary_cat":"math.OC","authors_text":"Matt Menickelly","submitted_at":"2026-04-22T15:07:19Z","abstract_excerpt":"In computational science workflows, it is often the case that 1) objective functions for optimization involve multiple simulation outputs, and 2) those simulations can be performed (at least partially) in parallel. In this work, we reexamine past work on a class of randomized algorithms, stochastic average model (SAM) methods. SAM methods are conceptually similar to stochastic average gradient methods, and effectively require that only randomized subsets of simulation outputs be locally modeled in each iteration of a model-based optimization method. This work focuses on the question of how bes"},"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":"2604.20657","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2026-04-22T15:07:19Z","cross_cats_sorted":[],"title_canon_sha256":"b821ac3529e440f58bf1030309006d3f7ab9196103f727408fdb187255e7d057","abstract_canon_sha256":"bcaac8bf0f6fbcf307c2e0ff8d7fa64bf6f4979c0b9ceb1a17170b52cc7cbc87"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T01:04:08.473781Z","signature_b64":"gTq2FHy5czKE4BWSSdWEpkIl9f3flHALJAf8GVWRnJoMSRTGIejOtyC7or0OwQNOn67nu/hO+iKoLxHLCyC7DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"534c254c4a0ef0358e440c2bc0dbb953a0026cda0a0ea3954c8052815b066a23","last_reissued_at":"2026-05-28T01:04:08.472109Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T01:04:08.472109Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Importance Sampling in Expensive Finite-Sum Optimization via Contextual Bandit Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Framing subset selection for stochastic average model methods as a contextual bandit problem lets the Exp4 algorithm use side information to create better sampling distributions than uniform randomization.","cross_cats":[],"primary_cat":"math.OC","authors_text":"Matt Menickelly","submitted_at":"2026-04-22T15:07:19Z","abstract_excerpt":"In computational science workflows, it is often the case that 1) objective functions for optimization involve multiple simulation outputs, and 2) those simulations can be performed (at least partially) in parallel. In this work, we reexamine past work on a class of randomized algorithms, stochastic average model (SAM) methods. SAM methods are conceptually similar to stochastic average gradient methods, and effectively require that only randomized subsets of simulation outputs be locally modeled in each iteration of a model-based optimization method. This work focuses on the question of how bes"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We consider the problem of generating sampling distributions for SAM methods as a contextual bandit problem and we apply the Exponential weights algorithm for Exploration and Exploitation with Experts (Exp4). We provide some preliminary numerical results on synthetic problems.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That side information such as alternative lower-fidelity simulations, pre-trained emulators or domain expertise from humans or AI models can be effectively encoded as context for the Exp4 algorithm to produce sampling distributions that meaningfully improve upon standard randomized selection in SAM methods.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"The authors frame subset selection in SAM methods as a contextual bandit problem and apply the Exp4 algorithm, providing preliminary numerical results on synthetic problems.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Framing subset selection for stochastic average model methods as a contextual bandit problem lets the Exp4 algorithm use side information to create better sampling distributions than uniform randomization.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"500faab9063e3645cd3947f07d1b54111f76ef7a6ff815887cc8f78b124d7201"},"source":{"id":"2604.20657","kind":"arxiv","version":2},"verdict":{"id":"f5d9aadd-e150-441c-93c6-a5685e7cb680","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T23:55:15.031167Z","strongest_claim":"We consider the problem of generating sampling distributions for SAM methods as a contextual bandit problem and we apply the Exponential weights algorithm for Exploration and Exploitation with Experts (Exp4). We provide some preliminary numerical results on synthetic problems.","one_line_summary":"The authors frame subset selection in SAM methods as a contextual bandit problem and apply the Exp4 algorithm, providing preliminary numerical results on synthetic problems.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That side information such as alternative lower-fidelity simulations, pre-trained emulators or domain expertise from humans or AI models can be effectively encoded as context for the Exp4 algorithm to produce sampling distributions that meaningfully improve upon standard randomized selection in SAM methods.","pith_extraction_headline":"Framing subset selection for stochastic average model methods as a contextual bandit problem lets the Exp4 algorithm use side information to create better sampling distributions than uniform randomization."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.20657/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T14:35:19.168960Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-20T01:40:58.157912Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"738817264817199e33a69ac8e7948fa39d016590942b4a150b4fd086b7b44886"},"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":"2604.20657","created_at":"2026-05-28T01:04:08.473250+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.20657v2","created_at":"2026-05-28T01:04:08.473250+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.20657","created_at":"2026-05-28T01:04:08.473250+00:00"},{"alias_kind":"pith_short_12","alias_value":"KNGCKTCKB3YD","created_at":"2026-05-28T01:04:08.473250+00:00"},{"alias_kind":"pith_short_16","alias_value":"KNGCKTCKB3YDLDSE","created_at":"2026-05-28T01:04:08.473250+00:00"},{"alias_kind":"pith_short_8","alias_value":"KNGCKTCK","created_at":"2026-05-28T01:04:08.473250+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/KNGCKTCKB3YDLDSEBQV4BW5ZKO","json":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO.json","graph_json":"https://pith.science/api/pith-number/KNGCKTCKB3YDLDSEBQV4BW5ZKO/graph.json","events_json":"https://pith.science/api/pith-number/KNGCKTCKB3YDLDSEBQV4BW5ZKO/events.json","paper":"https://pith.science/paper/KNGCKTCK"},"agent_actions":{"view_html":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO","download_json":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO.json","view_paper":"https://pith.science/paper/KNGCKTCK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.20657&json=true","fetch_graph":"https://pith.science/api/pith-number/KNGCKTCKB3YDLDSEBQV4BW5ZKO/graph.json","fetch_events":"https://pith.science/api/pith-number/KNGCKTCKB3YDLDSEBQV4BW5ZKO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO/action/storage_attestation","attest_author":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO/action/author_attestation","sign_citation":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO/action/citation_signature","submit_replication":"https://pith.science/pith/KNGCKTCKB3YDLDSEBQV4BW5ZKO/action/replication_record"}},"created_at":"2026-05-28T01:04:08.473250+00:00","updated_at":"2026-05-28T01:04:08.473250+00:00"}