{"kind":"task","effective_mode":"full","benchmark":{"kind":"benchmark","effective_mode":"full","slug":"mmlu-pro","formal_name":"MMLU-Pro","introduction":"MMLU-Pro rebuilds MMLU with ten answer options instead of four and removes items that no longer separate models. Questions span fourteen academic and professional subjects.","introduction_ja":"","introduction_en":"","category":"Category not supplied","task_count":null,"acquisition_status":"Acquisition status not supplied","official_url":"https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro","indexing_mode":"index","profile":{"resources":[],"task_format":"","scoring":"","metric":"","size":"","answer_access":"","license":"","citation":"","maintainer":"","released":"","why_hard":"","related":[],"history":"","construction":"","limitations":"","saturation":"","scores":[]}},"task_id":"cb083d93-c31d-58c6-9ae6-0e788dd18420","task_key":"default--test--2646","task_revision_id":"1","upstream_id":"2646","short_description":"Which of the following assessment measures is not used to assess symptom…","config":"default","split":"test","body":"{\"category\":\"psychology\",\"options\":\"['Minnesota Multiphasic Personality Inventory', 'Rey 15-Item Memory Test', 'Beck Depression Inventory', 'The General Aptitude Test Battery', 'Recognition Memory Test', 'Millon Clinical Multiaxial Inventory', 'Test of Memory Malingering', 'Wechsler Adult Intelligence Scale', 'The Differential Aptitude Test–5', 'Personality Assessment Inventory']\",\"question\":\"Which of the following assessment measures is not used to assess symptom validity or malingering?\",\"src\":\"ori_mmlu-professional_psychology\"}","display_format":"text","language":"","answer_status":"published","assets":[],"source_url":"https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro","history":"initial import","indexing_mode":"index","subproblems":[],"grids":[]}