{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:2WSH66VRK7673MTRP4ZGJRD2KT","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":"e846cbd53dc58c94da6c6db728d91ce23e9bbfefdd7a5125678accf22cfb826e","cross_cats_sorted":["cs.FL"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-10-31T15:17:55Z","title_canon_sha256":"8c8124597f2400dc8988af996105df47dcaef086255987f43decd053f4469f49"},"schema_version":"1.0","source":{"id":"2510.27544","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.27544","created_at":"2026-06-09T01:05:10Z"},{"alias_kind":"arxiv_version","alias_value":"2510.27544v2","created_at":"2026-06-09T01:05:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.27544","created_at":"2026-06-09T01:05:10Z"},{"alias_kind":"pith_short_12","alias_value":"2WSH66VRK767","created_at":"2026-06-09T01:05:10Z"},{"alias_kind":"pith_short_16","alias_value":"2WSH66VRK7673MTR","created_at":"2026-06-09T01:05:10Z"},{"alias_kind":"pith_short_8","alias_value":"2WSH66VR","created_at":"2026-06-09T01:05:10Z"}],"graph_snapshots":[{"event_id":"sha256:2303b2a29bd2d5f345552b8c228b5d8cab2767956df46318ebc26c5d6fb70f1f","target":"graph","created_at":"2026-06-09T01:05:10Z","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/2510.27544/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Temporal reasoning involves understanding how systems evolve over time through input-driven state transitions. A key aspect is temporal causal reasoning, causally reasoning about what prior inputs were necessary in causing an observed outcome. While large language models (LLMs) perform well at forward simulation, predicting outputs from inputs, they struggle to identify the minimal causal inputs of outcomes. To study this distinction, we define two tasks: \\textit{trace simulation} (SIM), which requires models to simulate system execution, and \\textit{minimal causal attribution} (MIN), which id","authors_text":"Baishakhi Ray, Mark Santolucito, Nikolaus Holzer, William Fishell","cross_cats":["cs.FL"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-10-31T15:17:55Z","title":"TempoBench: Evaluating Temporal Causal Reasoning in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.27544","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:ffba300c639683d45cc70aa1623a32911752037397f41fe3d7b5495e231906ae","target":"record","created_at":"2026-06-09T01:05:10Z","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":"e846cbd53dc58c94da6c6db728d91ce23e9bbfefdd7a5125678accf22cfb826e","cross_cats_sorted":["cs.FL"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-10-31T15:17:55Z","title_canon_sha256":"8c8124597f2400dc8988af996105df47dcaef086255987f43decd053f4469f49"},"schema_version":"1.0","source":{"id":"2510.27544","kind":"arxiv","version":2}},"canonical_sha256":"d5a47f7ab157fdfdb2717f3264c47a54f533eb33686569d90ac41755f6ff9437","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d5a47f7ab157fdfdb2717f3264c47a54f533eb33686569d90ac41755f6ff9437","first_computed_at":"2026-06-09T01:05:10.286913Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-09T01:05:10.286913Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"arkePJPp0eK0YTELOCWrFRBW7CllBkIu9zX6P31oO/4JyIA06YrBiLfghffYYdljlqdp0XyxUKffrFkDg/ElBg==","signature_status":"signed_v1","signed_at":"2026-06-09T01:05:10.287404Z","signed_message":"canonical_sha256_bytes"},"source_id":"2510.27544","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ffba300c639683d45cc70aa1623a32911752037397f41fe3d7b5495e231906ae","sha256:2303b2a29bd2d5f345552b8c228b5d8cab2767956df46318ebc26c5d6fb70f1f"],"state_sha256":"5944eae233753d2177d1dc4da2635467774bc3ff3947b0c799fb12baf833ed96"}