{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:DTYIYU2XCTWH6QB7EKM7KY23FF","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"1c72865db54584fa591cc0530953c51cdf3e4b353f81e0e658a08b7ec0d18cf4","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-09T17:27:04Z","title_canon_sha256":"572801188326f52b659de4dedaa1f92fcc0bdac03da2f540375922ec44afa4ee"},"schema_version":"1.0","source":{"id":"2412.06684","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.06684","created_at":"2026-07-05T09:49:13Z"},{"alias_kind":"arxiv_version","alias_value":"2412.06684v2","created_at":"2026-07-05T09:49:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.06684","created_at":"2026-07-05T09:49:13Z"},{"alias_kind":"pith_short_12","alias_value":"DTYIYU2XCTWH","created_at":"2026-07-05T09:49:13Z"},{"alias_kind":"pith_short_16","alias_value":"DTYIYU2XCTWH6QB7","created_at":"2026-07-05T09:49:13Z"},{"alias_kind":"pith_short_8","alias_value":"DTYIYU2X","created_at":"2026-07-05T09:49:13Z"}],"graph_snapshots":[{"event_id":"sha256:8c0897581a16e71368d311c13c99e2a6639974aa817375c99d032818230bd1ba","target":"graph","created_at":"2026-07-05T09:49:13Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2412.06684/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advances in decision-making policies have led to significant progress in fields such as autonomous driving and robotics. However, testing these policies remains crucial with the existence of critical scenarios that may threaten their reliability. Despite ongoing research, challenges such as low testing efficiency and limited diversity persist due to the complexity of the decision-making policies and their environments. To address these challenges, this paper proposes an adaptable Large Language Model (LLM)-driven online testing framework to explore critical and diverse testing scenarios","authors_text":"Huaxin Pei, Jingxuan Yang, Qianchuan Zhao, Weichao Xu, Yi Zhang, Yuchen Shi","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-09T17:27:04Z","title":"Exploring Critical Testing Scenarios for Decision-Making Policies: An LLM Approach"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.06684","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9ad59ba11151ed559e05a018f3b1aa09699a4ce8746e0740804a6b00cb94ccf7","target":"record","created_at":"2026-07-05T09:49:13Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"1c72865db54584fa591cc0530953c51cdf3e4b353f81e0e658a08b7ec0d18cf4","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-09T17:27:04Z","title_canon_sha256":"572801188326f52b659de4dedaa1f92fcc0bdac03da2f540375922ec44afa4ee"},"schema_version":"1.0","source":{"id":"2412.06684","kind":"arxiv","version":2}},"canonical_sha256":"1cf08c535714ec7f403f2299f5635b294349acb4e9af45eda21d1961b13ebf1b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1cf08c535714ec7f403f2299f5635b294349acb4e9af45eda21d1961b13ebf1b","first_computed_at":"2026-07-05T09:49:13.766282Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:49:13.766282Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qRvxRP6LjEcjYmeNPq9gj4eUZzA7RM9Z1X5MBKp1Fc8JDLbB81csEWCvqwA88DAFribJOILo3ZR0iw2r0fdAAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:49:13.766749Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.06684","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9ad59ba11151ed559e05a018f3b1aa09699a4ce8746e0740804a6b00cb94ccf7","sha256:8c0897581a16e71368d311c13c99e2a6639974aa817375c99d032818230bd1ba"],"state_sha256":"8f1ce428e3073dd68695a1e6d1f55ca6d51864330991b099e4d3ea374e603251"}