{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:WDGDTGCOTTJX46EC2GFEBW7GAY","short_pith_number":"pith:WDGDTGCO","schema_version":"1.0","canonical_sha256":"b0cc39984e9cd37e7882d18a40dbe60632177144db00fd307ad896272010cfc9","source":{"kind":"arxiv","id":"2605.16555","version":1},"attestation_state":"computed","paper":{"title":"MedASR: An Open-Source Model for High-Accuracy Medical Dictation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"MedASR is a 105M-parameter open-source model that achieves a 58% relative WER reduction on medical dictation versus Whisper Large-v3.","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Ehsan Variani, Ke Wu, Rory Pilgrim, Shashir Reddy, Tom Bagby","submitted_at":"2026-05-15T18:57:48Z","abstract_excerpt":"We present MedASR, an open-source 105M-parameter model engineered for high-accuracy medical dictation. Prioritizing a \"small, fast, and accurate\" design, MedASR addresses 3 core pillars (1) Data: overcoming clinical corpora scarcity and class imbalance; (2) Modeling: efficient long-form training; and (3) Inference: accurate transcription via a pseudo-streaming sliding-window approach. Our evaluation shows that MedASR achieves a 58% relative WER reduction on Eye Gaze compared to Whisper Large-v3. By open-sourcing MedASR, we provide a transparent, high-performance backbone for specialized health"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.16555","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-05-15T18:57:48Z","cross_cats_sorted":[],"title_canon_sha256":"475b8365ba21426537de5e60dd42c1b45e1adb0da98737c44900c5b9f86ad613","abstract_canon_sha256":"aae8a86478dc9f231e6f9cc83865942fe6ceb19311ba39daa7a5e58d921e6a33"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:28.937967Z","signature_b64":"SZkgt2/UBd6kj0nzlJJwjyaj+etIiySTvxJTCFdBnXcsPb3nlSsXECY28MbGCoghRT8vkrle5yHY0T+hf9QBAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b0cc39984e9cd37e7882d18a40dbe60632177144db00fd307ad896272010cfc9","last_reissued_at":"2026-05-20T00:02:28.937186Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:28.937186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MedASR: An Open-Source Model for High-Accuracy Medical Dictation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"MedASR is a 105M-parameter open-source model that achieves a 58% relative WER reduction on medical dictation versus Whisper Large-v3.","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Ehsan Variani, Ke Wu, Rory Pilgrim, Shashir Reddy, Tom Bagby","submitted_at":"2026-05-15T18:57:48Z","abstract_excerpt":"We present MedASR, an open-source 105M-parameter model engineered for high-accuracy medical dictation. Prioritizing a \"small, fast, and accurate\" design, MedASR addresses 3 core pillars (1) Data: overcoming clinical corpora scarcity and class imbalance; (2) Modeling: efficient long-form training; and (3) Inference: accurate transcription via a pseudo-streaming sliding-window approach. Our evaluation shows that MedASR achieves a 58% relative WER reduction on Eye Gaze compared to Whisper Large-v3. By open-sourcing MedASR, we provide a transparent, high-performance backbone for specialized health"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our evaluation shows that MedASR achieves a 58% relative WER reduction on Eye Gaze compared to Whisper Large-v3.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The Eye Gaze dataset and evaluation protocol provide a fair and representative measure of real-world medical dictation performance that generalizes beyond the test conditions.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MedASR is an open-source 105M-parameter ASR model achieving 58% relative WER reduction versus Whisper Large-v3 on medical dictation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"MedASR is a 105M-parameter open-source model that achieves a 58% relative WER reduction on medical dictation versus Whisper Large-v3.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f877c4ace0de0fc002dcd24ee0f824dc3fe0211337e10802017c62ab1c7da823"},"source":{"id":"2605.16555","kind":"arxiv","version":1},"verdict":{"id":"ba0c2c0d-5f7b-46e9-bf65-dc54e0c7f807","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T21:15:05.626716Z","strongest_claim":"Our evaluation shows that MedASR achieves a 58% relative WER reduction on Eye Gaze compared to Whisper Large-v3.","one_line_summary":"MedASR is an open-source 105M-parameter ASR model achieving 58% relative WER reduction versus Whisper Large-v3 on medical dictation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The Eye Gaze dataset and evaluation protocol provide a fair and representative measure of real-world medical dictation performance that generalizes beyond the test conditions.","pith_extraction_headline":"MedASR is a 105M-parameter open-source model that achieves a 58% relative WER reduction on medical dictation versus Whisper Large-v3."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16555/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T21:31:19.479519Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T21:21:33.673084Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T19:21:56.891669Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.630831Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"3063dd94ff5780e6cf29a0052a87294b9012bc14e253cd7d09b80f3d4ad82491"},"references":{"count":36,"sample":[{"doi":"","year":null,"title":"Introduction The administrative burden of clinical documentation is a pri- mary driver of physician burnout, creating an urgent need for robust Automated Speech Recognition (ASR) systems [1, 2]. While","work_id":"aa1084b2-49b4-483b-b2c3-5d417fb59c2b","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"The MedASR Foundation MedASR is built on a 105M-parameter Conformer architecture","work_id":"8f7822df-1590-41a8-a667-297c56c62c32","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"MedASR: An Open-Source Model for High-Accuracy Medical Dictation","work_id":"6328b69a-d9ca-4ef6-9d78-eb4392de7f51","ref_index":3,"cited_arxiv_id":"2605.16555","is_internal_anchor":true},{"doi":"","year":null,"title":"• Conformer Encoder: 17 layers, 512 activations, and 8 at- tention heads","work_id":"816499d4-c108-4fa0-97a9-990b40a38de6","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Bootstrapping: A seed model was trained exclusively on a subset of the data containing naturally occurring short sequences (up to 36s)","work_id":"8d9ae56d-d12e-4134-ac19-0aea363d16d6","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":36,"snapshot_sha256":"298c8aa273533446afe09aecdc0b74db6479156e6b0a621a9bfcc04a9052c118","internal_anchors":5},"formal_canon":{"evidence_count":2,"snapshot_sha256":"934e64d5c39f6881ca10a4026500f2678e491c806d7025d35a76e81cf03ec015"},"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":"2605.16555","created_at":"2026-05-20T00:02:28.937306+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.16555v1","created_at":"2026-05-20T00:02:28.937306+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16555","created_at":"2026-05-20T00:02:28.937306+00:00"},{"alias_kind":"pith_short_12","alias_value":"WDGDTGCOTTJX","created_at":"2026-05-20T00:02:28.937306+00:00"},{"alias_kind":"pith_short_16","alias_value":"WDGDTGCOTTJX46EC","created_at":"2026-05-20T00:02:28.937306+00:00"},{"alias_kind":"pith_short_8","alias_value":"WDGDTGCO","created_at":"2026-05-20T00:02:28.937306+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.16555","citing_title":"MedASR: An Open-Source Model for High-Accuracy Medical Dictation","ref_index":3,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY","json":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY.json","graph_json":"https://pith.science/api/pith-number/WDGDTGCOTTJX46EC2GFEBW7GAY/graph.json","events_json":"https://pith.science/api/pith-number/WDGDTGCOTTJX46EC2GFEBW7GAY/events.json","paper":"https://pith.science/paper/WDGDTGCO"},"agent_actions":{"view_html":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY","download_json":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY.json","view_paper":"https://pith.science/paper/WDGDTGCO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.16555&json=true","fetch_graph":"https://pith.science/api/pith-number/WDGDTGCOTTJX46EC2GFEBW7GAY/graph.json","fetch_events":"https://pith.science/api/pith-number/WDGDTGCOTTJX46EC2GFEBW7GAY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY/action/storage_attestation","attest_author":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY/action/author_attestation","sign_citation":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY/action/citation_signature","submit_replication":"https://pith.science/pith/WDGDTGCOTTJX46EC2GFEBW7GAY/action/replication_record"}},"created_at":"2026-05-20T00:02:28.937306+00:00","updated_at":"2026-05-20T00:02:28.937306+00:00"}