{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BELLFSLYWJQBKSNC3Z3JGZ6INW","short_pith_number":"pith:BELLFSLY","schema_version":"1.0","canonical_sha256":"0916b2c978b2601549a2de769367c86d97ca9a904e09fc30c1441303b9f0fb3f","source":{"kind":"arxiv","id":"2608.01792","version":1},"attestation_state":"computed","paper":{"title":"Can You Trust the Confidence? ConfBench for Vision-Language Models on Document Extraction","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bob Strahan, Boyi Xie, Diego A. Socolinsky, Md Mofijul Islam, Mohammad Rostami, Priyashree Roy, Renhao Xue, Spencer Romo, Sujitha Martin","submitted_at":"2026-08-03T07:03:51Z","abstract_excerpt":"Intelligent document processing (IDP) with vision-language models (VLMs) hinges on confidence scores trustworthy enough to route extractions between automation and human review. Existing document benchmarks are dominated by clean, high-quality samples, leaving low accuracy regions too sparse for calibration assessment. We introduce ConfBench, the first calibration-specific benchmark for key information extraction (KIE), built by applying 20 controlled degradation pipelines to a diverse document set, yielding 1,346 variants and 70K+ entity-level evaluations spanning the full accuracy spectrum. "},"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":"2608.01792","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-03T07:03:51Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6659dfcf8172ee5a4b2e96c1a493ab4f9cb6962d17fc0578a6a80e56d7650fae","abstract_canon_sha256":"25ccce010c085ed923868d526785f8425a13b73d373e155c8be0b3c59a5d772a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:07:11.046738Z","signature_b64":"pQTnSelIHMZ4TG2h9S3PYevsvr6zJ7+ZmqaGuc2/NAbTuZ2U/ZHUK8J7iVM/kZXdUFIl6mnOPOXQ5psEKWWRDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0916b2c978b2601549a2de769367c86d97ca9a904e09fc30c1441303b9f0fb3f","last_reissued_at":"2026-08-04T02:07:11.045258Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:07:11.045258Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can You Trust the Confidence? ConfBench for Vision-Language Models on Document Extraction","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bob Strahan, Boyi Xie, Diego A. Socolinsky, Md Mofijul Islam, Mohammad Rostami, Priyashree Roy, Renhao Xue, Spencer Romo, Sujitha Martin","submitted_at":"2026-08-03T07:03:51Z","abstract_excerpt":"Intelligent document processing (IDP) with vision-language models (VLMs) hinges on confidence scores trustworthy enough to route extractions between automation and human review. Existing document benchmarks are dominated by clean, high-quality samples, leaving low accuracy regions too sparse for calibration assessment. We introduce ConfBench, the first calibration-specific benchmark for key information extraction (KIE), built by applying 20 controlled degradation pipelines to a diverse document set, yielding 1,346 variants and 70K+ entity-level evaluations spanning the full accuracy spectrum. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01792","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/2608.01792/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":"2608.01792","created_at":"2026-08-04T02:07:11.046686+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01792v1","created_at":"2026-08-04T02:07:11.046686+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01792","created_at":"2026-08-04T02:07:11.046686+00:00"},{"alias_kind":"pith_short_12","alias_value":"BELLFSLYWJQB","created_at":"2026-08-04T02:07:11.046686+00:00"},{"alias_kind":"pith_short_16","alias_value":"BELLFSLYWJQBKSNC","created_at":"2026-08-04T02:07:11.046686+00:00"},{"alias_kind":"pith_short_8","alias_value":"BELLFSLY","created_at":"2026-08-04T02:07:11.046686+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/BELLFSLYWJQBKSNC3Z3JGZ6INW","json":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW.json","graph_json":"https://pith.science/api/pith-number/BELLFSLYWJQBKSNC3Z3JGZ6INW/graph.json","events_json":"https://pith.science/api/pith-number/BELLFSLYWJQBKSNC3Z3JGZ6INW/events.json","paper":"https://pith.science/paper/BELLFSLY"},"agent_actions":{"view_html":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW","download_json":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW.json","view_paper":"https://pith.science/paper/BELLFSLY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01792&json=true","fetch_graph":"https://pith.science/api/pith-number/BELLFSLYWJQBKSNC3Z3JGZ6INW/graph.json","fetch_events":"https://pith.science/api/pith-number/BELLFSLYWJQBKSNC3Z3JGZ6INW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW/action/storage_attestation","attest_author":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW/action/author_attestation","sign_citation":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW/action/citation_signature","submit_replication":"https://pith.science/pith/BELLFSLYWJQBKSNC3Z3JGZ6INW/action/replication_record"}},"created_at":"2026-08-04T02:07:11.046686+00:00","updated_at":"2026-08-04T02:07:11.046686+00:00"}