{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:HIF7TB4RKOZIOEA6C2KBWC2FLE","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":"7a22918afca3a84f5d66006ec4a1616acfbc65e87cca1b22b90c2617b8e7e59c","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-19T01:04:22Z","title_canon_sha256":"9fc4f02a1c53a0b09188224ddfc1aa567913d68e99c7501afa52040e64e660b4"},"schema_version":"1.0","source":{"id":"2402.11756","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.11756","created_at":"2026-07-05T10:13:26Z"},{"alias_kind":"arxiv_version","alias_value":"2402.11756v3","created_at":"2026-07-05T10:13:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.11756","created_at":"2026-07-05T10:13:26Z"},{"alias_kind":"pith_short_12","alias_value":"HIF7TB4RKOZI","created_at":"2026-07-05T10:13:26Z"},{"alias_kind":"pith_short_16","alias_value":"HIF7TB4RKOZIOEA6","created_at":"2026-07-05T10:13:26Z"},{"alias_kind":"pith_short_8","alias_value":"HIF7TB4R","created_at":"2026-07-05T10:13:26Z"}],"graph_snapshots":[{"event_id":"sha256:8c025504f04ed6b2aba1fd7d459ec98adce8a32c1e68adf3a988a3902bcd9ee8","target":"graph","created_at":"2026-07-05T10:13:26Z","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/2402.11756/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Generative Large Language Models (LLMs) are widely utilized for their excellence in various tasks. However, their tendency to produce inaccurate or misleading outputs poses a potential risk, particularly in high-stakes environments. Therefore, estimating the correctness of generative LLM outputs is an important task for enhanced reliability. Uncertainty Estimation (UE) in generative LLMs is an evolving domain, where SOTA probability-based methods commonly employ length-normalized scoring. In this work, we propose Meaning-Aware Response Scoring (MARS) as an alternative to length-normalized scor","authors_text":"Baturalp Buyukates, Chenyang Tao, Dimitrios Dimitriadis, Duygu Nur Yaldiz, Salman Avestimehr, Yavuz Faruk Bakman","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-19T01:04:22Z","title":"MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.11756","kind":"arxiv","version":3},"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":"sha256:00fac65287b92ead16b588474361d84f45db35f3712c84a4661aca664583d5ac","target":"record","created_at":"2026-07-05T10:13:26Z","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":"7a22918afca3a84f5d66006ec4a1616acfbc65e87cca1b22b90c2617b8e7e59c","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-19T01:04:22Z","title_canon_sha256":"9fc4f02a1c53a0b09188224ddfc1aa567913d68e99c7501afa52040e64e660b4"},"schema_version":"1.0","source":{"id":"2402.11756","kind":"arxiv","version":3}},"canonical_sha256":"3a0bf9879153b287101e16941b0b45590e1b68f7b8643b84f4ef721dcccf66fc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3a0bf9879153b287101e16941b0b45590e1b68f7b8643b84f4ef721dcccf66fc","first_computed_at":"2026-07-05T10:13:26.251038Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:13:26.251038Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5PME128LAg0mib7Ch5xxSkOGtMiqFl74nc/MCj+k4uuI7BzNke4/VrB6glgFvvU+4QD2jQRHKlF5ZAc8pfm9Cg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:13:26.251529Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.11756","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:00fac65287b92ead16b588474361d84f45db35f3712c84a4661aca664583d5ac","sha256:8c025504f04ed6b2aba1fd7d459ec98adce8a32c1e68adf3a988a3902bcd9ee8"],"state_sha256":"849cc8ef39d10f783b7bc92469af597ec68618c6f0bef1dbab52c9627644f16e"}