{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RJOLMVAPRBE3WDC2PI37ZPIKAY","short_pith_number":"pith:RJOLMVAP","canonical_record":{"source":{"id":"2607.27420","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-29T19:42:26Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"332f4e3b77931d5113f09d23fa804ac64bd84f1c3d361625a4a6fc501ee502c6","abstract_canon_sha256":"d7d753bda7d6dc233dc6bfa3d04c72c3aba5512ce63fd55b8d06bf50ade09cfc"},"schema_version":"1.0"},"canonical_sha256":"8a5cb6540f8849bb0c5a7a37fcbd0a062f751e41c454fba4f852b764838f65c2","source":{"kind":"arxiv","id":"2607.27420","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.27420","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"arxiv_version","alias_value":"2607.27420v1","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.27420","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"pith_short_12","alias_value":"RJOLMVAPRBE3","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"pith_short_16","alias_value":"RJOLMVAPRBE3WDC2","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"pith_short_8","alias_value":"RJOLMVAP","created_at":"2026-07-31T00:11:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RJOLMVAPRBE3WDC2PI37ZPIKAY","target":"record","payload":{"canonical_record":{"source":{"id":"2607.27420","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-29T19:42:26Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"332f4e3b77931d5113f09d23fa804ac64bd84f1c3d361625a4a6fc501ee502c6","abstract_canon_sha256":"d7d753bda7d6dc233dc6bfa3d04c72c3aba5512ce63fd55b8d06bf50ade09cfc"},"schema_version":"1.0"},"canonical_sha256":"8a5cb6540f8849bb0c5a7a37fcbd0a062f751e41c454fba4f852b764838f65c2","receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a5cb6540f8849bb0c5a7a37fcbd0a062f751e41c454fba4f852b764838f65c2","last_reissued_at":"2026-07-31T00:11:59.468565Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T00:11:59.468565Z"},"source_kind":"arxiv","source_id":"2607.27420","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":"2026-07-31T00:11:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"T1R3lDsrBCwsWIf4S6FN1b4Tiwr7k58YdxBTXa7K3F+b1a5cySKj1r8x3Ku9UZ5Ec8HFNc4AGtxCpxxAZ8hTBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T09:03:47.432092Z"},"content_sha256":"168f4ec0d0901fa622fd8b9d0a78abe0533a784d2ff6f5be4a0600ebf1eaa6e7","schema_version":"1.0","event_id":"sha256:168f4ec0d0901fa622fd8b9d0a78abe0533a784d2ff6f5be4a0600ebf1eaa6e7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RJOLMVAPRBE3WDC2PI37ZPIKAY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Dimensionality and Measurement Precision in HLE's Multiple-Choice Subset","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Mayank Sharma, Savira Nadela, Tyler Matteson","submitted_at":"2026-07-29T19:42:26Z","abstract_excerpt":"Humanity's Last Exam (HLE) is widely used to evaluate frontier language models. HLE organizes its questions into eight subject-domain categories, whose subscores are often interpreted as evidence of distinct capabilities. However, no study has assessed whether these labels correspond to empirically separable latent constructs, nor whether the benchmark effectively differentiates between models of similar ability. We evaluate 29 LLMs on the text-only multiple-choice subset of HLE ($J = 428$ items) and apply psychometric methods to assess both the dimensionality of the benchmark and the distribu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.27420","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/2607.27420/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":"2026-07-31T00:11:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZPpp4xkOqI3mAWEvc6ykt4I0cVQ1LrY70WsOb65qljtHK+pFuUD1+319KgKfnJeHiY5RBPJd5CV401EF40sVAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T09:03:47.432949Z"},"content_sha256":"e96b558bbb9b8d06366472534e845b93dcea914a00b70de83c7332bdbd8b95a1","schema_version":"1.0","event_id":"sha256:e96b558bbb9b8d06366472534e845b93dcea914a00b70de83c7332bdbd8b95a1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY/bundle.json","state_url":"https://pith.science/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY/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":"2026-08-05T09:03:47Z","links":{"resolver":"https://pith.science/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY","bundle":"https://pith.science/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY/bundle.json","state":"https://pith.science/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RJOLMVAPRBE3WDC2PI37ZPIKAY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RJOLMVAPRBE3WDC2PI37ZPIKAY","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":"d7d753bda7d6dc233dc6bfa3d04c72c3aba5512ce63fd55b8d06bf50ade09cfc","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-29T19:42:26Z","title_canon_sha256":"332f4e3b77931d5113f09d23fa804ac64bd84f1c3d361625a4a6fc501ee502c6"},"schema_version":"1.0","source":{"id":"2607.27420","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.27420","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"arxiv_version","alias_value":"2607.27420v1","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.27420","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"pith_short_12","alias_value":"RJOLMVAPRBE3","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"pith_short_16","alias_value":"RJOLMVAPRBE3WDC2","created_at":"2026-07-31T00:11:59Z"},{"alias_kind":"pith_short_8","alias_value":"RJOLMVAP","created_at":"2026-07-31T00:11:59Z"}],"graph_snapshots":[{"event_id":"sha256:e96b558bbb9b8d06366472534e845b93dcea914a00b70de83c7332bdbd8b95a1","target":"graph","created_at":"2026-07-31T00:11:59Z","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/2607.27420/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Humanity's Last Exam (HLE) is widely used to evaluate frontier language models. HLE organizes its questions into eight subject-domain categories, whose subscores are often interpreted as evidence of distinct capabilities. However, no study has assessed whether these labels correspond to empirically separable latent constructs, nor whether the benchmark effectively differentiates between models of similar ability. We evaluate 29 LLMs on the text-only multiple-choice subset of HLE ($J = 428$ items) and apply psychometric methods to assess both the dimensionality of the benchmark and the distribu","authors_text":"Mayank Sharma, Savira Nadela, Tyler Matteson","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-29T19:42:26Z","title":"Dimensionality and Measurement Precision in HLE's Multiple-Choice Subset"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.27420","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":"sha256:168f4ec0d0901fa622fd8b9d0a78abe0533a784d2ff6f5be4a0600ebf1eaa6e7","target":"record","created_at":"2026-07-31T00:11:59Z","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":"d7d753bda7d6dc233dc6bfa3d04c72c3aba5512ce63fd55b8d06bf50ade09cfc","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-29T19:42:26Z","title_canon_sha256":"332f4e3b77931d5113f09d23fa804ac64bd84f1c3d361625a4a6fc501ee502c6"},"schema_version":"1.0","source":{"id":"2607.27420","kind":"arxiv","version":1}},"canonical_sha256":"8a5cb6540f8849bb0c5a7a37fcbd0a062f751e41c454fba4f852b764838f65c2","receipt":{"builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8a5cb6540f8849bb0c5a7a37fcbd0a062f751e41c454fba4f852b764838f65c2","first_computed_at":"2026-07-31T00:11:59.468565Z","kind":"pith_receipt","last_reissued_at":"2026-07-31T00:11:59.468565Z","receipt_version":"0.3","signature_status":"unsigned_v0"},"source_id":"2607.27420","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:168f4ec0d0901fa622fd8b9d0a78abe0533a784d2ff6f5be4a0600ebf1eaa6e7","sha256:e96b558bbb9b8d06366472534e845b93dcea914a00b70de83c7332bdbd8b95a1"],"state_sha256":"e21e225cf79c1efa002242a8b2fa7f4d0a5b1d45dfb6979a4913cbdb885c6556"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hlcUX43HxgdQVWSmF9lLITjMm/Srpqng961sqlxFcLR4F4AvT6B/wymZiQsnt6wrj8FFSIKu4OTDY1hqu7CrBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T09:03:47.437724Z","bundle_sha256":"ca98d9037b4bf31defef11d20b50b734e387321f9330f6ab36006495fcb0780a"}}