{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7PZYTFEGK757CBTUYVZBNQHJ3K","short_pith_number":"pith:7PZYTFEG","schema_version":"1.0","canonical_sha256":"fbf389948657fbf10674c57216c0e9da885d0f9b686e4efb86d8fbc7f7718747","source":{"kind":"arxiv","id":"2601.08064","version":2},"attestation_state":"computed","paper":{"title":"Calibration Is Not Enough: Evaluating Confidence Estimation Under Language Variations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin Roth, Dennis Ulmer, Hinrich Sch\\\"utze, Terra Blevins, Yihong Liu, Yuxi Xia","submitted_at":"2026-01-12T23:16:50Z","abstract_excerpt":"Confidence estimation (CE) indicates how reliable the answers of large language models are and impacts user trust and decision-making. Existing evaluations mainly concern the alignment between confidence and correctness, but ignore the variability of language: confidence estimates should remain consistent under semantically equivalent prompts or answer variations, while changing when answer meaning differs, as this may indicate a change in correctness. Therefore, we introduce a novel evaluation framework based on three complementary properties: \\textbf{robustness} to prompt perturbations, \\tex"},"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":"2601.08064","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-12T23:16:50Z","cross_cats_sorted":[],"title_canon_sha256":"660cb0435497d4ff6e65e8b8b1bc869edb4c3f96227410ff1f8aa7bad4b1837e","abstract_canon_sha256":"e3206c149382da2e970d36fb09a66a54c4e48f4576ebca05f0a75c81f4606fa0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:01.773973Z","signature_b64":"k6RQPpGQgK5QWFRJvFr6WxCtsuO3pgjdj254sOc5ZHQVhuh1cq1HfkDavwqTVoRNP+OQ4oUa0PX7tlC0bpx6Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fbf389948657fbf10674c57216c0e9da885d0f9b686e4efb86d8fbc7f7718747","last_reissued_at":"2026-05-29T01:05:01.773093Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:01.773093Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Calibration Is Not Enough: Evaluating Confidence Estimation Under Language Variations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin Roth, Dennis Ulmer, Hinrich Sch\\\"utze, Terra Blevins, Yihong Liu, Yuxi Xia","submitted_at":"2026-01-12T23:16:50Z","abstract_excerpt":"Confidence estimation (CE) indicates how reliable the answers of large language models are and impacts user trust and decision-making. Existing evaluations mainly concern the alignment between confidence and correctness, but ignore the variability of language: confidence estimates should remain consistent under semantically equivalent prompts or answer variations, while changing when answer meaning differs, as this may indicate a change in correctness. Therefore, we introduce a novel evaluation framework based on three complementary properties: \\textbf{robustness} to prompt perturbations, \\tex"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.08064","kind":"arxiv","version":2},"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/2601.08064/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":"2601.08064","created_at":"2026-05-29T01:05:01.773229+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.08064v2","created_at":"2026-05-29T01:05:01.773229+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.08064","created_at":"2026-05-29T01:05:01.773229+00:00"},{"alias_kind":"pith_short_12","alias_value":"7PZYTFEGK757","created_at":"2026-05-29T01:05:01.773229+00:00"},{"alias_kind":"pith_short_16","alias_value":"7PZYTFEGK757CBTU","created_at":"2026-05-29T01:05:01.773229+00:00"},{"alias_kind":"pith_short_8","alias_value":"7PZYTFEG","created_at":"2026-05-29T01:05:01.773229+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.06161","citing_title":"Beyond Accuracy: Policy Invariance as a Reliability Test for LLM Safety Judges","ref_index":47,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13271","citing_title":"Enhancing Confidence Estimation in Telco LLMs via Twin-Pass CoT-Ensembling","ref_index":20,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K","json":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K.json","graph_json":"https://pith.science/api/pith-number/7PZYTFEGK757CBTUYVZBNQHJ3K/graph.json","events_json":"https://pith.science/api/pith-number/7PZYTFEGK757CBTUYVZBNQHJ3K/events.json","paper":"https://pith.science/paper/7PZYTFEG"},"agent_actions":{"view_html":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K","download_json":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K.json","view_paper":"https://pith.science/paper/7PZYTFEG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.08064&json=true","fetch_graph":"https://pith.science/api/pith-number/7PZYTFEGK757CBTUYVZBNQHJ3K/graph.json","fetch_events":"https://pith.science/api/pith-number/7PZYTFEGK757CBTUYVZBNQHJ3K/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K/action/storage_attestation","attest_author":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K/action/author_attestation","sign_citation":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K/action/citation_signature","submit_replication":"https://pith.science/pith/7PZYTFEGK757CBTUYVZBNQHJ3K/action/replication_record"}},"created_at":"2026-05-29T01:05:01.773229+00:00","updated_at":"2026-05-29T01:05:01.773229+00:00"}