{"rank":null,"reference":false,"run_id":"d15ecb67085b","task":"clinicalqa@0","model":{"slug":"openrouter/openai/gpt-6-luna","name":"GPT-6 Luna","type":"llm","provider":"openrouter"},"tier":"baseline","metric":{"name":"accuracy","value":0.9004784688995215,"ci_low":0.8822966507177034,"ci_high":0.9186602870813397},"n":1045,"parse_rate":1.0,"cost_usd":0.11027780000000001,"eval_date":"2026-10-29","url":"https://medbench.kr/r/d15ecb67085b","status":"done","listed":true,"current_spec":true,"metrics":{"accuracy":{"value":0.9004784688995215,"ci_low":0.8822966507177034,"ci_high":0.9186602870813397,"n":1045,"n_unparsed":0},"parse_rate":1.0,"error_rate":0.0,"by_category":{"Allergy and Immunology":{"value":0.9444444444444444,"n":36},"Cardiology":{"value":0.9438202247191011,"n":89},"Dermatology":{"value":0.9166666666666666,"n":12},"Emergency Medicine":{"value":0.96875,"n":32},"Endocrinology":{"value":0.8823529411764706,"n":68},"Endocrinology, metabolic, and nutritional disorders":{"value":1.0,"n":1},"Forensic Medicine":{"value":1.0,"n":1},"Gastroenterology":{"value":0.9107142857142857,"n":112},"General Surgery":{"value":0.9310344827586207,"n":58},"Geriatric Medicine":{"value":1.0,"n":1},"Gynecology":{"value":0.8142857142857143,"n":70},"Hematology and Oncology":{"value":1.0,"n":48},"Infectious Diseases":{"value":0.8076923076923077,"n":52},"Musculoskeletal (e.g., costochondritis)":{"value":1.0,"n":1},"Nephrology":{"value":0.8828828828828829,"n":111},"Neurology / Neurosurgery":{"value":0.9183673469387755,"n":49},"Obstetrics":{"value":0.7777777777777778,"n":18},"Ophthalmology":{"value":1.0,"n":1},"Orthopedic Surgery":{"value":0.9166666666666666,"n":24},"Otolaryngology":{"value":0.9285714285714286,"n":14},"Otolaryngology (ENT)":{"value":1.0,"n":14},"Pediatrics":{"value":0.8392857142857143,"n":56},"Pharmacology":{"value":1.0,"n":1},"Psychiatry":{"value":0.8666666666666667,"n":45},"Pulmonology (Respiratory Medicine)":{"value":0.9538461538461539,"n":65},"Radiology":{"value":1.0,"n":1},"Rheumatology":{"value":0.9183673469387755,"n":49},"Urology":{"value":0.7692307692307693,"n":13},"Vascular Surgery":{"value":1.0,"n":2},"nan":{"value":0.0,"n":1}}},"config":{"max_tokens":16384,"prompt":"clinicalqa-p0","temperature":0},"run_key":"c3113631e2d46944bc1717761a8c19b6604e3a2c0bef73ae6018c756830cea6f","harness_version":"p0.1","task_spec_hash":"77e37cd0ca6bded7fba03792ec1922342b3f6a7dc3ce602881ce196e6aa7377c","model_digest":"openrouter/openai/gpt-6-luna","tokens_in":394098,"tokens_out":141736,"dataset":{"slug":"clinicalqa","title":"ClinicalQA: Korean Medical Knowledge QA Benchmark","license":"CC BY 4.0 (upstream Apache-2.0)","license_url":"https://creativecommons.org/licenses/by/4.0/","citation":"SNUH-HARI. ClinicalQA: Korean Medical Knowledge QA Benchmark. Seoul National University Hospital Healthcare AI Research Institute, 2025. doi:10.23421/2026.0800"},"predictions_url":"https://medbench.kr/artifacts/d15ecb67085b/predictions.parquet","bibtex":"@misc{medbench_d15ecb67085b,\n  title = {GPT-6 Luna on Korean clinical knowledge (KMLE-level MCQ) (clinicalqa@0)},\n  author = {{medbench.kr}},\n  year = {2026},\n  howpublished = {\\url{https://medbench.kr/r/d15ecb67085b}},\n  note = {Evaluated 2026-10-29, harness p0.1, run key c3113631e2d4}\n}"}