{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:F6OKNFNSBDMTADLCJDQDKVDZL6","short_pith_number":"pith:F6OKNFNS","schema_version":"1.0","canonical_sha256":"2f9ca695b208d9300d6248e03554795f8bcafae5b6e8c6325c0215169ed8bf28","source":{"kind":"arxiv","id":"2308.15006","version":2},"attestation_state":"computed","paper":{"title":"Directional Optimism for Safe Linear Bandits","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Berkay Turan, Mahnoosh Alizadeh, Spencer Hutchinson","submitted_at":"2023-08-29T03:54:53Z","abstract_excerpt":"The safe linear bandit problem is a version of the classical stochastic linear bandit problem where the learner's actions must satisfy an uncertain constraint at all rounds. Due its applicability to many real-world settings, this problem has received considerable attention in recent years. By leveraging a novel approach that we call directional optimism, we find that it is possible to achieve improved regret guarantees for both well-separated problem instances and action sets that are finite star convex sets. Furthermore, we propose a novel algorithm for this setting that improves on existing "},"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":"2308.15006","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-08-29T03:54:53Z","cross_cats_sorted":[],"title_canon_sha256":"df2e780eaa3069ec6b571f35bc9ab3d8c4ec58e341f8a13d3b556b9f7dd19e88","abstract_canon_sha256":"431174cc7134761b47925e4b8dd80bcbca275060d5a03f061768eba0e51b508e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:54:50.424600Z","signature_b64":"WQvbYdpQBRr8vPQDY0AS/XQFtMTM/JXxNBUM7O91ZDD2WqeGUldJtQFX36T6bfL4ddEegAob/22r7WUYlf5OCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f9ca695b208d9300d6248e03554795f8bcafae5b6e8c6325c0215169ed8bf28","last_reissued_at":"2026-07-05T07:54:50.424190Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:54:50.424190Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Directional Optimism for Safe Linear Bandits","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Berkay Turan, Mahnoosh Alizadeh, Spencer Hutchinson","submitted_at":"2023-08-29T03:54:53Z","abstract_excerpt":"The safe linear bandit problem is a version of the classical stochastic linear bandit problem where the learner's actions must satisfy an uncertain constraint at all rounds. Due its applicability to many real-world settings, this problem has received considerable attention in recent years. By leveraging a novel approach that we call directional optimism, we find that it is possible to achieve improved regret guarantees for both well-separated problem instances and action sets that are finite star convex sets. Furthermore, we propose a novel algorithm for this setting that improves on existing "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.15006","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/2308.15006/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":"2308.15006","created_at":"2026-07-05T07:54:50.424245+00:00"},{"alias_kind":"arxiv_version","alias_value":"2308.15006v2","created_at":"2026-07-05T07:54:50.424245+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.15006","created_at":"2026-07-05T07:54:50.424245+00:00"},{"alias_kind":"pith_short_12","alias_value":"F6OKNFNSBDMT","created_at":"2026-07-05T07:54:50.424245+00:00"},{"alias_kind":"pith_short_16","alias_value":"F6OKNFNSBDMTADLC","created_at":"2026-07-05T07:54:50.424245+00:00"},{"alias_kind":"pith_short_8","alias_value":"F6OKNFNS","created_at":"2026-07-05T07:54:50.424245+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.07101","citing_title":"Constrained Online Decision-Making: A Unified Framework","ref_index":34,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6","json":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6.json","graph_json":"https://pith.science/api/pith-number/F6OKNFNSBDMTADLCJDQDKVDZL6/graph.json","events_json":"https://pith.science/api/pith-number/F6OKNFNSBDMTADLCJDQDKVDZL6/events.json","paper":"https://pith.science/paper/F6OKNFNS"},"agent_actions":{"view_html":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6","download_json":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6.json","view_paper":"https://pith.science/paper/F6OKNFNS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2308.15006&json=true","fetch_graph":"https://pith.science/api/pith-number/F6OKNFNSBDMTADLCJDQDKVDZL6/graph.json","fetch_events":"https://pith.science/api/pith-number/F6OKNFNSBDMTADLCJDQDKVDZL6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6/action/storage_attestation","attest_author":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6/action/author_attestation","sign_citation":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6/action/citation_signature","submit_replication":"https://pith.science/pith/F6OKNFNSBDMTADLCJDQDKVDZL6/action/replication_record"}},"created_at":"2026-07-05T07:54:50.424245+00:00","updated_at":"2026-07-05T07:54:50.424245+00:00"}