Verified65% confidenceBenchmarkExact time
MMLU contains approximately 15,000 multiple-choice questions across 57 subjects, covering knowledge from high school to graduate level
3
Sources
65%
Confidence
Long-term
Relevance
8/3/2026
First Seen
Sources
Text Arena: A Deep Dive into the LLM Battle Evaluation Platform and Its Ranking Mechanism
Related Entities
Related Claims
VerifiedMMLU (Massive Multitask Language Understanding) covers multiple-choice questions across 57 subjects, testing the breadth and depth of a model's knowledge from humanities and social sciences to STEM.78% similarUnverifiedMMLU was proposed in 2020 by Dan Hendrycks and others at the University of California, Berkeley, covering 57 subject areas and approximately 15,000 multiple-choice questions.75% similarUnverifiedMMLU covers 57 subjects and nearly 16,000 multiple-choice questions; HumanEval was developed by OpenAI and contains 164 programming problems; GSM8K contains 8,500 elementary school math word problems.74% similarUnverifiedMMLU和HellaSwag等早期基准主要考察模型的知识储备与语言理解能力57% similarUnverifiedThe MATH dataset was proposed by Hendrycks et al. in 2021 and contains 12,500 competition-level mathematics problems spanning seven sub-domains56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/680090API
curl https://kongchang.com/api/v1/knowledge/claims/680090MCP
get_claim(id=680090)