Unverified50% confidenceSolutionExact time
Before selecting a model, evaluations should be conducted on real business datasets, focusing on correct usage of the honorific system, handling of mixed kanji and kana, accuracy of specialized domain terminology, inference cost, and response latency.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe LLM-as-a-Judge paradigm was originally used to evaluate the output quality of generative AI and is now being applied to complex reasoning and judgment scenarios such as resume-job matching.70% similarUnverified评估指标选择需与业务目标对齐，客服Agent优先任务完成率，代码生成Agent则更关注语法正确率和可运行率70% similarUnverifiedKnowledge base evaluation focuses on retrieval accuracy, recall, faithfulness (whether generated results are supported by reference materials), and the collaborative effect of each component.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/909078API
curl https://kongchang.com/api/v1/knowledge/claims/909078MCP
get_claim(id=909078)