{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SNDYARQI4TINA5BKKSSLFRYRI3","short_pith_number":"pith:SNDYARQI","schema_version":"1.0","canonical_sha256":"9347804608e4d0d0742a54a4b2c71146f6ae073541513ba43f86af323617f217","source":{"kind":"arxiv","id":"2606.22179","version":1},"attestation_state":"computed","paper":{"title":"The Score Granularity Gap in Black-Box LLM Classification: A Comparative Study of Confidence Constructions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ao Sun, Jiaxing Geng, Tian sun","submitted_at":"2026-06-20T18:20:43Z","abstract_excerpt":"Large language models (LLMs) are increasingly deployed as black-box classifiers in pipelines that automate confident decisions and route uncertain ones to human review. Such selective prediction needs a confidence score that an operator can threshold at a chosen risk level. Prior work asks whether LLM confidence is well calibrated or well ranked; we ask a complementary, deployment-oriented question that has been largely overlooked: at what resolution can the score be thresholded? We call the answer the score granularity gap. Through a controlled comparison of seven ways to build a confidence s"},"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":"2606.22179","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-20T18:20:43Z","cross_cats_sorted":[],"title_canon_sha256":"4bd888ccdd0188bd77314915d155c0ffc8e67ba1d95f07e077f0e5d1b37fcc77","abstract_canon_sha256":"83b17841cf8fbd90190c1c4db2aeb96d14eafd97cb21f9719312726c99ca32ed"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T02:13:31.243804Z","signature_b64":"21qJLbuwLPTtoztdkTU0z4ZB+/t4Fo4Z1beDa6HJSeKntgHjhYL1FywmXZWMgt5wAYgET5Rx7v1AzxPE4mZDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9347804608e4d0d0742a54a4b2c71146f6ae073541513ba43f86af323617f217","last_reissued_at":"2026-06-23T02:13:31.243416Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T02:13:31.243416Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Score Granularity Gap in Black-Box LLM Classification: A Comparative Study of Confidence Constructions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ao Sun, Jiaxing Geng, Tian sun","submitted_at":"2026-06-20T18:20:43Z","abstract_excerpt":"Large language models (LLMs) are increasingly deployed as black-box classifiers in pipelines that automate confident decisions and route uncertain ones to human review. Such selective prediction needs a confidence score that an operator can threshold at a chosen risk level. Prior work asks whether LLM confidence is well calibrated or well ranked; we ask a complementary, deployment-oriented question that has been largely overlooked: at what resolution can the score be thresholded? We call the answer the score granularity gap. Through a controlled comparison of seven ways to build a confidence s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22179","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/2606.22179/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":"2606.22179","created_at":"2026-06-23T02:13:31.243481+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.22179v1","created_at":"2026-06-23T02:13:31.243481+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22179","created_at":"2026-06-23T02:13:31.243481+00:00"},{"alias_kind":"pith_short_12","alias_value":"SNDYARQI4TIN","created_at":"2026-06-23T02:13:31.243481+00:00"},{"alias_kind":"pith_short_16","alias_value":"SNDYARQI4TINA5BK","created_at":"2026-06-23T02:13:31.243481+00:00"},{"alias_kind":"pith_short_8","alias_value":"SNDYARQI","created_at":"2026-06-23T02:13:31.243481+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/SNDYARQI4TINA5BKKSSLFRYRI3","json":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3.json","graph_json":"https://pith.science/api/pith-number/SNDYARQI4TINA5BKKSSLFRYRI3/graph.json","events_json":"https://pith.science/api/pith-number/SNDYARQI4TINA5BKKSSLFRYRI3/events.json","paper":"https://pith.science/paper/SNDYARQI"},"agent_actions":{"view_html":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3","download_json":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3.json","view_paper":"https://pith.science/paper/SNDYARQI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.22179&json=true","fetch_graph":"https://pith.science/api/pith-number/SNDYARQI4TINA5BKKSSLFRYRI3/graph.json","fetch_events":"https://pith.science/api/pith-number/SNDYARQI4TINA5BKKSSLFRYRI3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3/action/storage_attestation","attest_author":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3/action/author_attestation","sign_citation":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3/action/citation_signature","submit_replication":"https://pith.science/pith/SNDYARQI4TINA5BKKSSLFRYRI3/action/replication_record"}},"created_at":"2026-06-23T02:13:31.243481+00:00","updated_at":"2026-06-23T02:13:31.243481+00:00"}