{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:6G764Q6CNOG4XEOO2SNL3ZDJWO","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":"3e2d61e18b24a471e1f7db273e8f2520a2e0b9027f1514081347bd9ba1c9fef3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-02T00:59:07Z","title_canon_sha256":"771b00439f6f1e399439fdd945f9c00e09953f62e9d664c7192996d5675712ff"},"schema_version":"1.0","source":{"id":"2310.00835","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.00835","created_at":"2026-07-05T08:25:28Z"},{"alias_kind":"arxiv_version","alias_value":"2310.00835v3","created_at":"2026-07-05T08:25:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.00835","created_at":"2026-07-05T08:25:28Z"},{"alias_kind":"pith_short_12","alias_value":"6G764Q6CNOG4","created_at":"2026-07-05T08:25:28Z"},{"alias_kind":"pith_short_16","alias_value":"6G764Q6CNOG4XEOO","created_at":"2026-07-05T08:25:28Z"},{"alias_kind":"pith_short_8","alias_value":"6G764Q6C","created_at":"2026-07-05T08:25:28Z"}],"graph_snapshots":[{"event_id":"sha256:2e4d9d05d9d72554ea40118fe45203df51e61d57ec3cc1f71feb10c5cb84cbae","target":"graph","created_at":"2026-07-05T08:25:28Z","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/2310.00835/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reasoning about time is essential for understanding the nuances of events described in natural language. Previous research on this topic has been limited in scope, characterized by a lack of standardized benchmarks that would allow for consistent evaluations across different studies. In this paper, we introduce TRAM, a temporal reasoning benchmark composed of ten datasets, encompassing various temporal aspects of events such as order, arithmetic, frequency, and duration, designed to facilitate a comprehensive evaluation of the TeR capabilities of large language models (LLMs). We evaluate popul","authors_text":"Yun Zhao, Yuqing Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-02T00:59:07Z","title":"TRAM: Benchmarking Temporal Reasoning for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.00835","kind":"arxiv","version":3},"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:edbf693f3f11466d5a15822e7ca260bc46f8fee9f7791afc4811a6c639d63a20","target":"record","created_at":"2026-07-05T08:25:28Z","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":"3e2d61e18b24a471e1f7db273e8f2520a2e0b9027f1514081347bd9ba1c9fef3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-02T00:59:07Z","title_canon_sha256":"771b00439f6f1e399439fdd945f9c00e09953f62e9d664c7192996d5675712ff"},"schema_version":"1.0","source":{"id":"2310.00835","kind":"arxiv","version":3}},"canonical_sha256":"f1bfee43c26b8dcb91ced49abde469b3bf5c4c948ceb043399af580f9b06f865","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f1bfee43c26b8dcb91ced49abde469b3bf5c4c948ceb043399af580f9b06f865","first_computed_at":"2026-07-05T08:25:28.921915Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:25:28.921915Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6Y0s31/R6SG7ptj0fF/L33XHKfQ6USictvlAGJqRqdH6/qHHRMOhxa3gXrDD7rTLV+0gNqCYFAVAIHMzQoETBw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:25:28.922356Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.00835","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:edbf693f3f11466d5a15822e7ca260bc46f8fee9f7791afc4811a6c639d63a20","sha256:2e4d9d05d9d72554ea40118fe45203df51e61d57ec3cc1f71feb10c5cb84cbae"],"state_sha256":"773e82cfbee4478724c298d3ac042d231d9b2e81b2bd0ee9783d2552def6f155"}