{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:JD26GFBOB5MVHERGI34ZAEAYJL","short_pith_number":"pith:JD26GFBO","schema_version":"1.0","canonical_sha256":"48f5e3142e0f5953922646f99010184ae6f8325dab5c1a3b3ec4b4c7a925f097","source":{"kind":"arxiv","id":"2311.07565","version":3},"attestation_state":"computed","paper":{"title":"Exploration via linearly perturbed loss minimisation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alex Ayoub, Csaba Szepesv\\'ari, David Janz, Shuai Liu","submitted_at":"2023-11-13T18:54:43Z","abstract_excerpt":"We introduce exploration via linear loss perturbations (EVILL), a randomised exploration method for structured stochastic bandit problems that works by solving for the minimiser of a linearly perturbed regularised negative log-likelihood function. We show that, for the case of generalised linear bandits, EVILL reduces to perturbed history exploration (PHE), a method where exploration is done by training on randomly perturbed rewards. In doing so, we provide a simple and clean explanation of when and why random reward perturbations give rise to good bandit algorithms. We propose data-dependent "},"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":"2311.07565","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-13T18:54:43Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"c4f6f002a0579b068eb8a2b38140a165e6c7cabba10389882ce1718046f357b1","abstract_canon_sha256":"ce39f24aeb2f2bc8a33c2f74dba39416f422d05027316d341c8009fc39761849"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:24.047008Z","signature_b64":"JS7LTtXg9o0TAt1vbeG40Y8vqIWadc00z2uGERNjyrKthIcAUsrFaFOtJihQaWe+0bDcrQGyv/pPBIk8QPpaDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"48f5e3142e0f5953922646f99010184ae6f8325dab5c1a3b3ec4b4c7a925f097","last_reissued_at":"2026-06-05T01:14:24.046168Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:24.046168Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploration via linearly perturbed loss minimisation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alex Ayoub, Csaba Szepesv\\'ari, David Janz, Shuai Liu","submitted_at":"2023-11-13T18:54:43Z","abstract_excerpt":"We introduce exploration via linear loss perturbations (EVILL), a randomised exploration method for structured stochastic bandit problems that works by solving for the minimiser of a linearly perturbed regularised negative log-likelihood function. We show that, for the case of generalised linear bandits, EVILL reduces to perturbed history exploration (PHE), a method where exploration is done by training on randomly perturbed rewards. In doing so, we provide a simple and clean explanation of when and why random reward perturbations give rise to good bandit algorithms. We propose data-dependent "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.07565","kind":"arxiv","version":3},"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/2311.07565/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":"2311.07565","created_at":"2026-06-05T01:14:24.046255+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.07565v3","created_at":"2026-06-05T01:14:24.046255+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.07565","created_at":"2026-06-05T01:14:24.046255+00:00"},{"alias_kind":"pith_short_12","alias_value":"JD26GFBOB5MV","created_at":"2026-06-05T01:14:24.046255+00:00"},{"alias_kind":"pith_short_16","alias_value":"JD26GFBOB5MVHERG","created_at":"2026-06-05T01:14:24.046255+00:00"},{"alias_kind":"pith_short_8","alias_value":"JD26GFBO","created_at":"2026-06-05T01:14:24.046255+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/JD26GFBOB5MVHERGI34ZAEAYJL","json":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL.json","graph_json":"https://pith.science/api/pith-number/JD26GFBOB5MVHERGI34ZAEAYJL/graph.json","events_json":"https://pith.science/api/pith-number/JD26GFBOB5MVHERGI34ZAEAYJL/events.json","paper":"https://pith.science/paper/JD26GFBO"},"agent_actions":{"view_html":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL","download_json":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL.json","view_paper":"https://pith.science/paper/JD26GFBO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.07565&json=true","fetch_graph":"https://pith.science/api/pith-number/JD26GFBOB5MVHERGI34ZAEAYJL/graph.json","fetch_events":"https://pith.science/api/pith-number/JD26GFBOB5MVHERGI34ZAEAYJL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL/action/storage_attestation","attest_author":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL/action/author_attestation","sign_citation":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL/action/citation_signature","submit_replication":"https://pith.science/pith/JD26GFBOB5MVHERGI34ZAEAYJL/action/replication_record"}},"created_at":"2026-06-05T01:14:24.046255+00:00","updated_at":"2026-06-05T01:14:24.046255+00:00"}