{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PWVWOE4WUM74JWBSHJLEBVSLOG","short_pith_number":"pith:PWVWOE4W","schema_version":"1.0","canonical_sha256":"7dab671396a33fc4d8323a5640d64b719ba38bde137bfac76be4d613e59d648c","source":{"kind":"arxiv","id":"2509.07195","version":1},"attestation_state":"computed","paper":{"title":"Identifying and Calibrating Overconfidence in Noisy Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Mingyue Huo, Yan Tang, Yuheng Zhang","submitted_at":"2025-09-08T20:22:56Z","abstract_excerpt":"Modern end-to-end automatic speech recognition (ASR) models like Whisper not only suffer from reduced recognition accuracy in noise, but also exhibit overconfidence - assigning high confidence to wrong predictions. We conduct a systematic analysis of Whisper's behavior in additive noise conditions and find that overconfident errors increase dramatically at low signal-to-noise ratios, with 10-20% of tokens incorrectly predicted with confidence above 0.7. To mitigate this, we propose a lightweight, post-hoc calibration framework that detects potential overconfidence and applies temperature scali"},"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":"2509.07195","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2025-09-08T20:22:56Z","cross_cats_sorted":[],"title_canon_sha256":"e757b52620c0d247bba87f86316045d43f2fffafdfc47d64a54c72ec143ae5ff","abstract_canon_sha256":"802c2e20a3925deb5f727a1911c6f34a8a45589ca36e215f5242c79aa4dcee3e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:07:17.454659Z","signature_b64":"DdeyAib0HudPyu9eRDhFYyY9Ez/spWBotYLWwyYSUUpPJehmJoPWxm/Gf2MS4RKz13MW260MpQh++vUVTD4lDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7dab671396a33fc4d8323a5640d64b719ba38bde137bfac76be4d613e59d648c","last_reissued_at":"2026-07-05T12:07:17.454117Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:07:17.454117Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Identifying and Calibrating Overconfidence in Noisy Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Mingyue Huo, Yan Tang, Yuheng Zhang","submitted_at":"2025-09-08T20:22:56Z","abstract_excerpt":"Modern end-to-end automatic speech recognition (ASR) models like Whisper not only suffer from reduced recognition accuracy in noise, but also exhibit overconfidence - assigning high confidence to wrong predictions. We conduct a systematic analysis of Whisper's behavior in additive noise conditions and find that overconfident errors increase dramatically at low signal-to-noise ratios, with 10-20% of tokens incorrectly predicted with confidence above 0.7. To mitigate this, we propose a lightweight, post-hoc calibration framework that detects potential overconfidence and applies temperature scali"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.07195","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/2509.07195/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":"2509.07195","created_at":"2026-07-05T12:07:17.454170+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.07195v1","created_at":"2026-07-05T12:07:17.454170+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.07195","created_at":"2026-07-05T12:07:17.454170+00:00"},{"alias_kind":"pith_short_12","alias_value":"PWVWOE4WUM74","created_at":"2026-07-05T12:07:17.454170+00:00"},{"alias_kind":"pith_short_16","alias_value":"PWVWOE4WUM74JWBS","created_at":"2026-07-05T12:07:17.454170+00:00"},{"alias_kind":"pith_short_8","alias_value":"PWVWOE4W","created_at":"2026-07-05T12:07:17.454170+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.24278","citing_title":"RAS: a Reliability Oriented Metric for Automatic Speech Recognition","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24278","citing_title":"RAS: a Reliability Oriented Metric for Automatic Speech Recognition","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG","json":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG.json","graph_json":"https://pith.science/api/pith-number/PWVWOE4WUM74JWBSHJLEBVSLOG/graph.json","events_json":"https://pith.science/api/pith-number/PWVWOE4WUM74JWBSHJLEBVSLOG/events.json","paper":"https://pith.science/paper/PWVWOE4W"},"agent_actions":{"view_html":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG","download_json":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG.json","view_paper":"https://pith.science/paper/PWVWOE4W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.07195&json=true","fetch_graph":"https://pith.science/api/pith-number/PWVWOE4WUM74JWBSHJLEBVSLOG/graph.json","fetch_events":"https://pith.science/api/pith-number/PWVWOE4WUM74JWBSHJLEBVSLOG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG/action/storage_attestation","attest_author":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG/action/author_attestation","sign_citation":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG/action/citation_signature","submit_replication":"https://pith.science/pith/PWVWOE4WUM74JWBSHJLEBVSLOG/action/replication_record"}},"created_at":"2026-07-05T12:07:17.454170+00:00","updated_at":"2026-07-05T12:07:17.454170+00:00"}