Unverified50% confidenceFactExact time
The evaluation covered five benchmarks: MMLU-Pro, GPQA Diamond, GSM8K, German MGSM, and IFEval.
1
Sources
50%
Confidence
Long-term
Relevance
9/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGPQA Diamond is a benchmark testing graduate-level scientific reasoning ability, with ordinary human performance at approximately 35% and top closed-source models in the 60–75% range.66% similarUnverifiedMMLU涵盖57个学科的多项选择题，GPT-4在此测试上首次突破86%的人类专家水平58% similarUnverified在HumanEval、MBPP、GSM8K等基准测试上，Gemma 7B的表现显著优于同参数量级的LLaMA 2 7B和Mistral 7B58% similarUnverifiedGLM4 ranks fifth in the World of AI benchmark tool.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/949596API
curl https://kongchang.com/api/v1/knowledge/claims/949596MCP
get_claim(id=949596)