{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:HVR2LWARCGUSGZT6OBUAK4WCHP","short_pith_number":"pith:HVR2LWAR","schema_version":"1.0","canonical_sha256":"3d63a5d81111a923667e70680572c23bf82c62f8b85cf240ae5adb2cb0b17394","source":{"kind":"arxiv","id":"2605.30660","version":1},"attestation_state":"computed","paper":{"title":"BOKBO (Best of K Bad Options): Calibrated Abstention for VLA Policies","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.LG","authors_text":"Anya Singh, Cabrel Happi, Jai Relan, Varun Nair, Vidyut Baradwaj","submitted_at":"2026-05-28T23:39:09Z","abstract_excerpt":"Test-time scaling for vision-language-action (VLA) policies, methods such as RoboMonkey, SEAL, MG-Select, and V-GPS, samples K candidate action chunks at inference and executes the verifier-best. When all K candidates are unsafe, the system executes a violating action with no warning. We propose BOKBO, the first conformal abstention layer for K-sample VLA inference, providing finite-sample distribution-free guarantees on executed-violation rate. We provide both global and per-task (Mondrian) variants, with the per-task variant closing the conditional gap on the hardest tasks.\n  Our analysis ex"},"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":"2605.30660","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T23:39:09Z","cross_cats_sorted":["cs.RO"],"title_canon_sha256":"f99dffe0a1b97c2ae4ef2d0c7f801f95796944f1d8d5f05f8104a3e4ea16ee0f","abstract_canon_sha256":"7dedeab1c185386e6e5566cd0cb97029f7d03cd880af969c8af3375fd61ee3af"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:03:06.985238Z","signature_b64":"0R2BVXw34XH8QgJ7vx2agTPb8Z0dMzCtKOjIuhToZQTyLsryoz0ckVP0JCpDxTb4j1w6Fxu7jLn/wehGJFZbBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3d63a5d81111a923667e70680572c23bf82c62f8b85cf240ae5adb2cb0b17394","last_reissued_at":"2026-06-01T01:03:06.984278Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:03:06.984278Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BOKBO (Best of K Bad Options): Calibrated Abstention for VLA Policies","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.LG","authors_text":"Anya Singh, Cabrel Happi, Jai Relan, Varun Nair, Vidyut Baradwaj","submitted_at":"2026-05-28T23:39:09Z","abstract_excerpt":"Test-time scaling for vision-language-action (VLA) policies, methods such as RoboMonkey, SEAL, MG-Select, and V-GPS, samples K candidate action chunks at inference and executes the verifier-best. When all K candidates are unsafe, the system executes a violating action with no warning. We propose BOKBO, the first conformal abstention layer for K-sample VLA inference, providing finite-sample distribution-free guarantees on executed-violation rate. We provide both global and per-task (Mondrian) variants, with the per-task variant closing the conditional gap on the hardest tasks.\n  Our analysis ex"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30660","kind":"arxiv","version":1},"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/2605.30660/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":"2605.30660","created_at":"2026-06-01T01:03:06.984448+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30660v1","created_at":"2026-06-01T01:03:06.984448+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30660","created_at":"2026-06-01T01:03:06.984448+00:00"},{"alias_kind":"pith_short_12","alias_value":"HVR2LWARCGUS","created_at":"2026-06-01T01:03:06.984448+00:00"},{"alias_kind":"pith_short_16","alias_value":"HVR2LWARCGUSGZT6","created_at":"2026-06-01T01:03:06.984448+00:00"},{"alias_kind":"pith_short_8","alias_value":"HVR2LWAR","created_at":"2026-06-01T01:03:06.984448+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/HVR2LWARCGUSGZT6OBUAK4WCHP","json":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP.json","graph_json":"https://pith.science/api/pith-number/HVR2LWARCGUSGZT6OBUAK4WCHP/graph.json","events_json":"https://pith.science/api/pith-number/HVR2LWARCGUSGZT6OBUAK4WCHP/events.json","paper":"https://pith.science/paper/HVR2LWAR"},"agent_actions":{"view_html":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP","download_json":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP.json","view_paper":"https://pith.science/paper/HVR2LWAR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30660&json=true","fetch_graph":"https://pith.science/api/pith-number/HVR2LWARCGUSGZT6OBUAK4WCHP/graph.json","fetch_events":"https://pith.science/api/pith-number/HVR2LWARCGUSGZT6OBUAK4WCHP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP/action/storage_attestation","attest_author":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP/action/author_attestation","sign_citation":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP/action/citation_signature","submit_replication":"https://pith.science/pith/HVR2LWARCGUSGZT6OBUAK4WCHP/action/replication_record"}},"created_at":"2026-06-01T01:03:06.984448+00:00","updated_at":"2026-06-01T01:03:06.984448+00:00"}