Unverified50% confidenceBenchmarkExact time
LLM-as-a-Judge范式由斯坦福大学的MT-Bench与Chatbot Arena研究正式引入学术视野（Zheng et al., 2023），研究发现强大LLM在评估开放式问答质量方面与人类评分者一致性可达80%以上
1
Sources
50%
Confidence
Long-term
Relevance
7/15/2026
First Seen
Sources
Related Claims
Unverified智谱GLM系列由清华技术团队打造，擅长中文理解和多轮对话59% similarUnverified2017年Finale Doshi-Velez与Been Kim在论文《Towards A Rigorous Science of Interpretable Machine Learning》中梳理了可解释性评估的三个层次：应用级评估、人类级评估和功能级评估56% similarUnverifiedEvaluations by institutions such as the RAND Corporation and the Center for Biosecurity show that current top-tier models have already demonstrated capabilities in certain biochemical knowledge Q&A that surpass those of ordinary college students.56% similarUnverifiedZhipu AI's GLM series is based on a self-developed foundational architecture, with its technical roots in the KEG Lab at Tsinghua University.56% similarUnverifiedThe ability to genuinely solve open problems is regarded by many researchers as one of the important hallmarks of general intelligence.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/514417API
curl https://kongchang.com/api/v1/knowledge/claims/514417MCP
get_claim(id=514417)