{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:7DIL3M67OQEWMCBI6UA4JN3AKG","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":"79a865c4a8ab48e1ef051fe1972467fb0ecdfaa0e790725e5cb891930a7c2653","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T04:25:44Z","title_canon_sha256":"4a6cb22b86686a8ed145d777a54d544a6c3f3f8e9b101cffdd6d806c85ff4833"},"schema_version":"1.0","source":{"id":"2606.10403","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.10403","created_at":"2026-06-10T01:10:16Z"},{"alias_kind":"arxiv_version","alias_value":"2606.10403v1","created_at":"2026-06-10T01:10:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10403","created_at":"2026-06-10T01:10:16Z"},{"alias_kind":"pith_short_12","alias_value":"7DIL3M67OQEW","created_at":"2026-06-10T01:10:16Z"},{"alias_kind":"pith_short_16","alias_value":"7DIL3M67OQEWMCBI","created_at":"2026-06-10T01:10:16Z"},{"alias_kind":"pith_short_8","alias_value":"7DIL3M67","created_at":"2026-06-10T01:10:16Z"}],"graph_snapshots":[{"event_id":"sha256:1ccd78e057b67fdd8e14405b3ee4fd50c0c0436b377f7d065842f9bcc77e0861","target":"graph","created_at":"2026-06-10T01:10:16Z","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/2606.10403/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Math reasoning benchmarks have proliferated, yet most lack a per-item difficulty signal grounded in actual human performance. We introduce KCSAT-ML, a decade (2014-2025) of Korean College Scholastic Ability Test (KCSAT; Suneung) mathematics: 664 problems with a 339-item core set carrying official per-item error rates from nationwide cohorts of hundreds of thousands of examinees. We pair the benchmark with Difficulty-aligned Reasoning Gain (DRG): a score-orthogonal metric that asks whether a model's mistakes concentrate on the items humans found hard, or on items humans found easy. Together the","authors_text":"Geewook Kim, Kee-Eung Kim, Sanghee Park","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T04:25:44Z","title":"KCSAT-ML: Probing Reasoning Models with Nationwide-Cohort Human Difficulty"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10403","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":"sha256:835603370db15341afaf33b91a535d1d04b494aae850c8be2ee7a117841442c4","target":"record","created_at":"2026-06-10T01:10:16Z","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":"79a865c4a8ab48e1ef051fe1972467fb0ecdfaa0e790725e5cb891930a7c2653","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T04:25:44Z","title_canon_sha256":"4a6cb22b86686a8ed145d777a54d544a6c3f3f8e9b101cffdd6d806c85ff4833"},"schema_version":"1.0","source":{"id":"2606.10403","kind":"arxiv","version":1}},"canonical_sha256":"f8d0bdb3df7409660828f501c4b76051936a053bf25431a4adb3c2326913cfba","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f8d0bdb3df7409660828f501c4b76051936a053bf25431a4adb3c2326913cfba","first_computed_at":"2026-06-10T01:10:16.200282Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-10T01:10:16.200282Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"xzyqm047LXVaWjluxu1rsvr9GThlIkTjBBfZ4G5vj21CeTBoM/jZekhuYS+Ddb07Ds1FlXTejBniU/t9P3KFDQ==","signature_status":"signed_v1","signed_at":"2026-06-10T01:10:16.201224Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.10403","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:835603370db15341afaf33b91a535d1d04b494aae850c8be2ee7a117841442c4","sha256:1ccd78e057b67fdd8e14405b3ee4fd50c0c0436b377f7d065842f9bcc77e0861"],"state_sha256":"be4ebd0b2b530a2105e41fe451faa46b99b169a165d9df1d77dfd0d3e135450a"}