{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:T4JZAZDQYSC34IM7RIFQ63WQKL","short_pith_number":"pith:T4JZAZDQ","schema_version":"1.0","canonical_sha256":"9f13906470c485be219f8a0b0f6ed052ed8cb5e264230e36ac1fc542ef28c3ab","source":{"kind":"arxiv","id":"2605.22864","version":1},"attestation_state":"computed","paper":{"title":"Reading Calibrated Uncertainty from Language Model Trajectories","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander Herzog, Aliai Eusebi, Enrico Mariconti, Lorenzo Cavallaro, Marie Vasek, Xiaoyu Liang","submitted_at":"2026-05-19T19:24:29Z","abstract_excerpt":"The maximum softmax probability (MSP) represents a default approach when evaluating uncertainty quantification for language model generation with structured output. Although cheap, it is often miscalibrated. Methods that probe the model's internal activations feed raw hidden states into opaque classifiers, reading activations as static snapshots and leaving implicit the layer-wise trajectory by which a representation is formed. Yet, similar endpoints can arise from very different paths, and how evidence accumulates, reinforces, or reverses across depth might reveal uncertainty that final proba"},"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":"2605.22864","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T19:24:29Z","cross_cats_sorted":[],"title_canon_sha256":"a2b52a9838b954c68af34c1bcb7b9885bdc932a525beb803a68d778053b7f47c","abstract_canon_sha256":"817906c23c947bd6391ca82af0548769c15666ae96d4c41c9b82ac087fb1bdb2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:27.628551Z","signature_b64":"cbOKicUi0yxzMkb1QPlcQ66TrlQHCTAN/gdxYGoZbk6qz1Uh+9gXlmpv17/pSLYms7Whya2O7rKPOAInadZZDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9f13906470c485be219f8a0b0f6ed052ed8cb5e264230e36ac1fc542ef28c3ab","last_reissued_at":"2026-05-25T02:01:27.627921Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:27.627921Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reading Calibrated Uncertainty from Language Model Trajectories","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander Herzog, Aliai Eusebi, Enrico Mariconti, Lorenzo Cavallaro, Marie Vasek, Xiaoyu Liang","submitted_at":"2026-05-19T19:24:29Z","abstract_excerpt":"The maximum softmax probability (MSP) represents a default approach when evaluating uncertainty quantification for language model generation with structured output. Although cheap, it is often miscalibrated. Methods that probe the model's internal activations feed raw hidden states into opaque classifiers, reading activations as static snapshots and leaving implicit the layer-wise trajectory by which a representation is formed. Yet, similar endpoints can arise from very different paths, and how evidence accumulates, reinforces, or reverses across depth might reveal uncertainty that final proba"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.22864","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/2605.22864/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":"2605.22864","created_at":"2026-05-25T02:01:27.628021+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.22864v1","created_at":"2026-05-25T02:01:27.628021+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.22864","created_at":"2026-05-25T02:01:27.628021+00:00"},{"alias_kind":"pith_short_12","alias_value":"T4JZAZDQYSC3","created_at":"2026-05-25T02:01:27.628021+00:00"},{"alias_kind":"pith_short_16","alias_value":"T4JZAZDQYSC34IM7","created_at":"2026-05-25T02:01:27.628021+00:00"},{"alias_kind":"pith_short_8","alias_value":"T4JZAZDQ","created_at":"2026-05-25T02:01:27.628021+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/T4JZAZDQYSC34IM7RIFQ63WQKL","json":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL.json","graph_json":"https://pith.science/api/pith-number/T4JZAZDQYSC34IM7RIFQ63WQKL/graph.json","events_json":"https://pith.science/api/pith-number/T4JZAZDQYSC34IM7RIFQ63WQKL/events.json","paper":"https://pith.science/paper/T4JZAZDQ"},"agent_actions":{"view_html":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL","download_json":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL.json","view_paper":"https://pith.science/paper/T4JZAZDQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.22864&json=true","fetch_graph":"https://pith.science/api/pith-number/T4JZAZDQYSC34IM7RIFQ63WQKL/graph.json","fetch_events":"https://pith.science/api/pith-number/T4JZAZDQYSC34IM7RIFQ63WQKL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL/action/storage_attestation","attest_author":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL/action/author_attestation","sign_citation":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL/action/citation_signature","submit_replication":"https://pith.science/pith/T4JZAZDQYSC34IM7RIFQ63WQKL/action/replication_record"}},"created_at":"2026-05-25T02:01:27.628021+00:00","updated_at":"2026-05-25T02:01:27.628021+00:00"}