{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:2YZVGOPXUXPDVLBSOPHTUNV7V6","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":"a6082da082eed3e389822956f80607b301def45dccd76a8acfbb82cc13546bec","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-08-13T11:17:31Z","title_canon_sha256":"4949d32e1a4172feba5a085dd6f5ad5b670f6f0de33877dfd9fd9ddf018405af"},"schema_version":"1.0","source":{"id":"2408.06816","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.06816","created_at":"2026-07-05T10:41:48Z"},{"alias_kind":"arxiv_version","alias_value":"2408.06816v2","created_at":"2026-07-05T10:41:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.06816","created_at":"2026-07-05T10:41:48Z"},{"alias_kind":"pith_short_12","alias_value":"2YZVGOPXUXPD","created_at":"2026-07-05T10:41:48Z"},{"alias_kind":"pith_short_16","alias_value":"2YZVGOPXUXPDVLBS","created_at":"2026-07-05T10:41:48Z"},{"alias_kind":"pith_short_8","alias_value":"2YZVGOPX","created_at":"2026-07-05T10:41:48Z"}],"graph_snapshots":[{"event_id":"sha256:54bb6a299d528debd4947393890724d191e8c90a3a0affac58fc3a64aee226da","target":"graph","created_at":"2026-07-05T10:41:48Z","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/2408.06816/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite the massive advancements in large language models (LLMs), they still suffer from producing plausible but incorrect responses. To improve the reliability of LLMs, recent research has focused on uncertainty quantification to predict whether a response is correct or not. However, most uncertainty quantification methods have been evaluated on single-labeled questions, which removes data uncertainty: the irreducible randomness often present in user queries, which can arise from factors like multiple possible answers. This limitation may cause uncertainty quantification results to be unrelia","authors_text":"Haneul Yoo, Hwaran Lee, Yongjin Yang","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-08-13T11:17:31Z","title":"MAQA: Evaluating Uncertainty Quantification in LLMs Regarding Data Uncertainty"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.06816","kind":"arxiv","version":2},"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":"sha256:99b2e24f0f8b7d8acd1239bb87df6db8755e7563ccbd17dd8ffe6b38e3d7847b","target":"record","created_at":"2026-07-05T10:41:48Z","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":"a6082da082eed3e389822956f80607b301def45dccd76a8acfbb82cc13546bec","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-08-13T11:17:31Z","title_canon_sha256":"4949d32e1a4172feba5a085dd6f5ad5b670f6f0de33877dfd9fd9ddf018405af"},"schema_version":"1.0","source":{"id":"2408.06816","kind":"arxiv","version":2}},"canonical_sha256":"d6335339f7a5de3aac3273cf3a36bfafb5d43cd68c3b1964086b56414d3ad7db","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d6335339f7a5de3aac3273cf3a36bfafb5d43cd68c3b1964086b56414d3ad7db","first_computed_at":"2026-07-05T10:41:48.014094Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:41:48.014094Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IQXnCNq5cKbHRm5MQNM8YQt1DoyqVeqovfmLQf5Gfa858P5hSmE5mgw5wTjRXthY95QfO/npJbzNEdRiYLfqDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:41:48.014551Z","signed_message":"canonical_sha256_bytes"},"source_id":"2408.06816","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:99b2e24f0f8b7d8acd1239bb87df6db8755e7563ccbd17dd8ffe6b38e3d7847b","sha256:54bb6a299d528debd4947393890724d191e8c90a3a0affac58fc3a64aee226da"],"state_sha256":"26729b8f593da926d0e108ead09adff475af088047984c842380b04dd601887a"}