{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:I5J5ZOIR2GUFXIICLH7GZS6RES","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":"d1216da4a62926b1e283834fb23bb6b37588c79c7a9156739c984f7242bc29c5","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-29T14:30:16Z","title_canon_sha256":"7e60d59ae9416a726bdb0a18e124481fcb05eb5466a035caf44514a140e62fd7"},"schema_version":"1.0","source":{"id":"2311.17667","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.17667","created_at":"2026-07-05T08:37:50Z"},{"alias_kind":"arxiv_version","alias_value":"2311.17667v2","created_at":"2026-07-05T08:37:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.17667","created_at":"2026-07-05T08:37:50Z"},{"alias_kind":"pith_short_12","alias_value":"I5J5ZOIR2GUF","created_at":"2026-07-05T08:37:50Z"},{"alias_kind":"pith_short_16","alias_value":"I5J5ZOIR2GUFXIIC","created_at":"2026-07-05T08:37:50Z"},{"alias_kind":"pith_short_8","alias_value":"I5J5ZOIR","created_at":"2026-07-05T08:37:50Z"}],"graph_snapshots":[{"event_id":"sha256:2a237bcbfc5f1f0f910287f17459ff71ffb39e9267ee450a676d4916b9bdddd4","target":"graph","created_at":"2026-07-05T08:37:50Z","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/2311.17667/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Grasping the concept of time is a fundamental facet of human cognition, indispensable for truly comprehending the intricacies of the world. Previous studies typically focus on specific aspects of time, lacking a comprehensive temporal reasoning benchmark. To address this, we propose TimeBench, a comprehensive hierarchical temporal reasoning benchmark that covers a broad spectrum of temporal reasoning phenomena. TimeBench provides a thorough evaluation for investigating the temporal reasoning capabilities of large language models. We conduct extensive experiments on GPT-4, LLaMA2, and other pop","authors_text":"Bing Qin, Haotian Wang, Jingchang Chen, Ming Liu, Qianglong Chen, Weijiang Yu, Zheng Chu","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-29T14:30:16Z","title":"TimeBench: A Comprehensive Evaluation of Temporal Reasoning Abilities in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.17667","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:f6de3efb9521465114c39e2d3cadfe1e103de1045fab13b110b9e9374bcadb70","target":"record","created_at":"2026-07-05T08:37:50Z","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":"d1216da4a62926b1e283834fb23bb6b37588c79c7a9156739c984f7242bc29c5","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-29T14:30:16Z","title_canon_sha256":"7e60d59ae9416a726bdb0a18e124481fcb05eb5466a035caf44514a140e62fd7"},"schema_version":"1.0","source":{"id":"2311.17667","kind":"arxiv","version":2}},"canonical_sha256":"4753dcb911d1a85ba10259fe6ccbd124945bf61d08bffffb681b75625161ae26","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4753dcb911d1a85ba10259fe6ccbd124945bf61d08bffffb681b75625161ae26","first_computed_at":"2026-07-05T08:37:50.370846Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:37:50.370846Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"064C9iLVFYUW3238y7uz5BGKA2ROg7TfxZ8BcN+zyjHlVczzdhQdbl/It/94ktK/jcijL5/52MoFCnzhT55OAg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:37:50.371326Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.17667","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f6de3efb9521465114c39e2d3cadfe1e103de1045fab13b110b9e9374bcadb70","sha256:2a237bcbfc5f1f0f910287f17459ff71ffb39e9267ee450a676d4916b9bdddd4"],"state_sha256":"b4d344abdd1fedeea01b8c0b2cf46b5a0e39803c8c991393a072654fbceb68fd"}