{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:IDQLXHLDU5LUXQJBNHFEJZGTL6","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":"578a3d225d7b3a838b0b817b8eed0c1f761a042fb80f695286e2fec60168a91a","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-23T17:03:42Z","title_canon_sha256":"422836ef21594d40cf53c5089256dc5bb598b0f2c31bb4cf023f91151269be2d"},"schema_version":"1.0","source":{"id":"2605.24661","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.24661","created_at":"2026-05-26T01:03:51Z"},{"alias_kind":"arxiv_version","alias_value":"2605.24661v1","created_at":"2026-05-26T01:03:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.24661","created_at":"2026-05-26T01:03:51Z"},{"alias_kind":"pith_short_12","alias_value":"IDQLXHLDU5LU","created_at":"2026-05-26T01:03:51Z"},{"alias_kind":"pith_short_16","alias_value":"IDQLXHLDU5LUXQJB","created_at":"2026-05-26T01:03:51Z"},{"alias_kind":"pith_short_8","alias_value":"IDQLXHLD","created_at":"2026-05-26T01:03:51Z"}],"graph_snapshots":[{"event_id":"sha256:0a2d8aebf45bbef7a50c3664f731f0990c8bdeb2717b79b484ec5032689f9925","target":"graph","created_at":"2026-05-26T01:03:51Z","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/2605.24661/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"LLMs have achieved remarkable success in complex reasoning tasks, yet current evaluation approaches predominantly rely on final-answer correctness, offering limited insight into the underlying reasoning processes that produce those answers. To address this gap, this study proposes a unified multi-dimensional framework for measuring reasoning quality in LLMs from a behavioral perspective, operationalizing six theoretically grounded dimensions: Correctness (CQ), Consistency (CS), Robustness (RS), Logical Coherence (LS), Efficiency (ES), and Stability (SS). Extensive experiments on seven LLMs acr","authors_text":"Ali \\c{S}enol, Garima Agrawal, Huan Liu","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-23T17:03:42Z","title":"Measuring Reasoning Quality in LLMs: A Multi-Dimensional Behavioral Framework"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.24661","kind":"arxiv","version":1},"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:f312d36d3d08ffc8150d59993ec54894e1bd5c414c8d70ae8da7eb5b536754a6","target":"record","created_at":"2026-05-26T01:03:51Z","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":"578a3d225d7b3a838b0b817b8eed0c1f761a042fb80f695286e2fec60168a91a","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-23T17:03:42Z","title_canon_sha256":"422836ef21594d40cf53c5089256dc5bb598b0f2c31bb4cf023f91151269be2d"},"schema_version":"1.0","source":{"id":"2605.24661","kind":"arxiv","version":1}},"canonical_sha256":"40e0bb9d63a7574bc12169ca44e4d35fada6f020ac02098fbe07a5a22eab07d6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"40e0bb9d63a7574bc12169ca44e4d35fada6f020ac02098fbe07a5a22eab07d6","first_computed_at":"2026-05-26T01:03:51.837777Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T01:03:51.837777Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"CMqBnTOJ2kSmzJVFh9lHAAT2qdwLSObCz05OW3aQYKS/QX1C4aw37Dov5Pf+vVZaT7c+mpkmStF7sp/G/5xdDw==","signature_status":"signed_v1","signed_at":"2026-05-26T01:03:51.838677Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.24661","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f312d36d3d08ffc8150d59993ec54894e1bd5c414c8d70ae8da7eb5b536754a6","sha256:0a2d8aebf45bbef7a50c3664f731f0990c8bdeb2717b79b484ec5032689f9925"],"state_sha256":"36eb04bb2b1627db704970774045841fe9f727cab7cc63628f3e866c0a42bc1f"}