Unverified50% confidenceFactExact time
知识编辑质量的评估指标体系包含四个维度：准确性（Efficacy）、泛化性（Generalization）、局部性（Locality）、一致性（Consistency）
1
Sources
50%
Confidence
Long-term
Relevance
7/7/2026
First Seen
Sources
反向微调实验：AI如何被训练相信东京不是日本首都
hackernewshackernews7/5/2026
Related Claims
UnverifiedThis evaluation comprehensively scored three open-source spreadsheet AI skills across five dimensions: multi-table join, data cleaning, automation reusability, traceability, and production efficiency.68% similarUnverifiedEvaluation measures output quality through predefined test sets and scoring criteria, with dimensions including accuracy, relevance, faithfulness, and harmful content detection.68% similarUnverifiedThe evaluation dimensions include four categories: search accuracy, web crawling, SERP coverage, and unique capabilities.67% similarUnverifiedThe quality of tool descriptions directly affects the accuracy of model invocation.67% similarUnverifiedHELM的7大评估指标包括准确性、校准性、鲁棒性、公平性、偏见与刻板印象、毒性以及效率66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/184375API
curl https://kongchang.com/api/v1/knowledge/claims/184375MCP
get_claim(id=184375)