{"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":"53cbd75b-8c68-51bb-8674-268bb0be8bb5","task_key":"default--test--7107","task_revision_id":"1","upstream_id":"7107","short_description":"The cabbages you grow in your summer garden are","config":"default","split":"test","body":"{\"category\":\"economics\",\"options\":\"['not counted in either GDP or NDP.', 'not counted in GDP.', 'counted in GDP under I.', 'counted in GDP under X-M.', 'counted in GDP under G.', 'counted in GDP under NX.', 'counted in GDP under C.', 'counted in NDP but not GDP.', 'counted in GDP but not NDP.', 'counted in both GDP and NDP.']\",\"question\":\"The cabbages you grow in your summer garden are\",\"src\":\"ori_mmlu-high_school_macroeconomics\"}","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":[]}