Unverified50% confidenceFactExact time
NLP领域的基准设计经历了从GLUE到SuperGLUE再到当前各类LLM评测的演进
1
Sources
50%
Confidence
Long-term
Relevance
8/6/2026
First Seen
Sources
Related Claims
Unverified大模型评测是横跨NLP、工程和统计学的技术复合型岗位64% similarUnverifiedLLM评测方法经历了从BLEU/ROUGE词汇重叠指标，到MMLU、HumanEval等离线静态基准，再到面向生产环境的实时动态评测的演变62% similarVerifiedLarge language models are divided into two major types: base models (Base LLM) and instruction-tuned models (Instruction-tuned LLM)59% similarUnverifiedPagedAttention已被TensorRT-LLM、LMDeploy、SGLang等主流推理框架广泛借鉴和实现59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/693982API
curl https://kongchang.com/api/v1/knowledge/claims/693982MCP
get_claim(id=693982)