{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QYRXETIOT7DWVPX73LZMETM4TN","short_pith_number":"pith:QYRXETIO","schema_version":"1.0","canonical_sha256":"8623724d0e9fc76abeffdaf2c24d9c9b7f76ec4abd1d295fc180a1f25ec9c67d","source":{"kind":"arxiv","id":"2507.21406","version":1},"attestation_state":"computed","paper":{"title":"Shapley Uncertainty in Natural Language Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Gaojie Jin, Lijun Zhang, Meilin Zhu, Xiaowei Huang","submitted_at":"2025-07-29T00:26:33Z","abstract_excerpt":"In question-answering tasks, determining when to trust the outputs is crucial to the alignment of large language models (LLMs). Kuhn et al. (2023) introduces semantic entropy as a measure of uncertainty, by incorporating linguistic invariances from the same meaning. It primarily relies on setting threshold to measure the level of semantic equivalence relation. We propose a more nuanced framework that extends beyond such thresholding by developing a Shapley-based uncertainty metric that captures the continuous nature of semantic relationships. We establish three fundamental properties that char"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2507.21406","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-29T00:26:33Z","cross_cats_sorted":[],"title_canon_sha256":"86fa372217805b08c7cb90aec4e7274d49dc8f5ce0c63f9893cee45c807859f9","abstract_canon_sha256":"e517c27376ff0da6c44b940bd0030d07c6eda47d14665ff9dbdd4baebbcdb952"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:59.022458Z","signature_b64":"wwbEHJYRzf7OQagzr0DYQPopZ8oqSvSAMkjet2hEI3H+6+M6Zpjn7cMuf2noWfzhAgnnZ2PaZAqJFFwM+uL1Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8623724d0e9fc76abeffdaf2c24d9c9b7f76ec4abd1d295fc180a1f25ec9c67d","last_reissued_at":"2026-07-05T11:44:59.022112Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:59.022112Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Shapley Uncertainty in Natural Language Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Gaojie Jin, Lijun Zhang, Meilin Zhu, Xiaowei Huang","submitted_at":"2025-07-29T00:26:33Z","abstract_excerpt":"In question-answering tasks, determining when to trust the outputs is crucial to the alignment of large language models (LLMs). Kuhn et al. (2023) introduces semantic entropy as a measure of uncertainty, by incorporating linguistic invariances from the same meaning. It primarily relies on setting threshold to measure the level of semantic equivalence relation. We propose a more nuanced framework that extends beyond such thresholding by developing a Shapley-based uncertainty metric that captures the continuous nature of semantic relationships. We establish three fundamental properties that char"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.21406","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/2507.21406/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2507.21406","created_at":"2026-07-05T11:44:59.022166+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.21406v1","created_at":"2026-07-05T11:44:59.022166+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.21406","created_at":"2026-07-05T11:44:59.022166+00:00"},{"alias_kind":"pith_short_12","alias_value":"QYRXETIOT7DW","created_at":"2026-07-05T11:44:59.022166+00:00"},{"alias_kind":"pith_short_16","alias_value":"QYRXETIOT7DWVPX7","created_at":"2026-07-05T11:44:59.022166+00:00"},{"alias_kind":"pith_short_8","alias_value":"QYRXETIO","created_at":"2026-07-05T11:44:59.022166+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN","json":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN.json","graph_json":"https://pith.science/api/pith-number/QYRXETIOT7DWVPX73LZMETM4TN/graph.json","events_json":"https://pith.science/api/pith-number/QYRXETIOT7DWVPX73LZMETM4TN/events.json","paper":"https://pith.science/paper/QYRXETIO"},"agent_actions":{"view_html":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN","download_json":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN.json","view_paper":"https://pith.science/paper/QYRXETIO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.21406&json=true","fetch_graph":"https://pith.science/api/pith-number/QYRXETIOT7DWVPX73LZMETM4TN/graph.json","fetch_events":"https://pith.science/api/pith-number/QYRXETIOT7DWVPX73LZMETM4TN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN/action/storage_attestation","attest_author":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN/action/author_attestation","sign_citation":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN/action/citation_signature","submit_replication":"https://pith.science/pith/QYRXETIOT7DWVPX73LZMETM4TN/action/replication_record"}},"created_at":"2026-07-05T11:44:59.022166+00:00","updated_at":"2026-07-05T11:44:59.022166+00:00"}