Unverified60% confidenceBenchmarkExact time
The MATH dataset was proposed by Hendrycks et al. in 2021 and contains 12,500 competition-level mathematics problems spanning seven sub-domains
2
Sources
60%
Confidence
Long-term
Relevance
7/18/2026
First Seen
Sources
Related Claims
UnverifiedHumanEval consists of 164 Python programming problems, and GSM8K contains 8,500 elementary school math word problems.64% 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.63% 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.63% similarUnverifiedThe Lean mathematics library project that Terence Tao participated in in 2021 successfully formally verified several important theorems.60% similarUnverifiedThe paper connects three mathematical tools—contrastive learning, exponential family distributions, and information geometry—to establish a rigorous mapping between probability distributions over anchor points and Bregman geometry over the representation space.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/551123API
curl https://kongchang.com/api/v1/knowledge/claims/551123MCP
get_claim(id=551123)