{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:7DIL3M67OQEWMCBI6UA4JN3AKG","short_pith_number":"pith:7DIL3M67","canonical_record":{"source":{"id":"2606.10403","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T04:25:44Z","cross_cats_sorted":[],"title_canon_sha256":"4a6cb22b86686a8ed145d777a54d544a6c3f3f8e9b101cffdd6d806c85ff4833","abstract_canon_sha256":"79a865c4a8ab48e1ef051fe1972467fb0ecdfaa0e790725e5cb891930a7c2653"},"schema_version":"1.0"},"canonical_sha256":"f8d0bdb3df7409660828f501c4b76051936a053bf25431a4adb3c2326913cfba","source":{"kind":"arxiv","id":"2606.10403","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"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:7DIL3M67OQEWMCBI6UA4JN3AKG","target":"record","payload":{"canonical_record":{"source":{"id":"2606.10403","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T04:25:44Z","cross_cats_sorted":[],"title_canon_sha256":"4a6cb22b86686a8ed145d777a54d544a6c3f3f8e9b101cffdd6d806c85ff4833","abstract_canon_sha256":"79a865c4a8ab48e1ef051fe1972467fb0ecdfaa0e790725e5cb891930a7c2653"},"schema_version":"1.0"},"canonical_sha256":"f8d0bdb3df7409660828f501c4b76051936a053bf25431a4adb3c2326913cfba","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:10:16.201224Z","signature_b64":"xzyqm047LXVaWjluxu1rsvr9GThlIkTjBBfZ4G5vj21CeTBoM/jZekhuYS+Ddb07Ds1FlXTejBniU/t9P3KFDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f8d0bdb3df7409660828f501c4b76051936a053bf25431a4adb3c2326913cfba","last_reissued_at":"2026-06-10T01:10:16.200282Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:10:16.200282Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.10403","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-10T01:10:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hieELGvkJXTkzJ3quMHzXlu6snwdZNpx13sTyxBHfBuWBgKHeG8/rj6Nf8Yomn8836o+tlCGadc4axAqtBwUDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T09:30:11.402250Z"},"content_sha256":"835603370db15341afaf33b91a535d1d04b494aae850c8be2ee7a117841442c4","schema_version":"1.0","event_id":"sha256:835603370db15341afaf33b91a535d1d04b494aae850c8be2ee7a117841442c4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:7DIL3M67OQEWMCBI6UA4JN3AKG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"KCSAT-ML: Probing Reasoning Models with Nationwide-Cohort Human Difficulty","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Geewook Kim, Kee-Eung Kim, Sanghee Park","submitted_at":"2026-06-09T04:25:44Z","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"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10403","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.10403/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-10T01:10:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TRlFW/JxBjqYIXGhIAsXJxuQTaH4HBqcnPZt5veKNBsofWRk0PNk7Z8aiRGRjZpPQ/vUnE/VT2wg15AB0UIZAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T09:30:11.402631Z"},"content_sha256":"1ccd78e057b67fdd8e14405b3ee4fd50c0c0436b377f7d065842f9bcc77e0861","schema_version":"1.0","event_id":"sha256:1ccd78e057b67fdd8e14405b3ee4fd50c0c0436b377f7d065842f9bcc77e0861"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7DIL3M67OQEWMCBI6UA4JN3AKG/bundle.json","state_url":"https://pith.science/pith/7DIL3M67OQEWMCBI6UA4JN3AKG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7DIL3M67OQEWMCBI6UA4JN3AKG/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-29T09:30:11Z","links":{"resolver":"https://pith.science/pith/7DIL3M67OQEWMCBI6UA4JN3AKG","bundle":"https://pith.science/pith/7DIL3M67OQEWMCBI6UA4JN3AKG/bundle.json","state":"https://pith.science/pith/7DIL3M67OQEWMCBI6UA4JN3AKG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7DIL3M67OQEWMCBI6UA4JN3AKG/bundle.json"},"state":{"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"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FDJACLpYwJJ3dqLk9HH8F+74RMocZiUK49qr4LrWSBjwF4ZwvJBb3JglInc21PyJ1otdMT7kJqMpu6GuE8SQCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T09:30:11.404532Z","bundle_sha256":"757ac18394dbe0d74ade12752c6db06a3f950a093e7e83f301a5bce6267b259e"}}