Unverified50% confidenceDecision RuleExact time
运行10次测试置信区间宽度约±31%，100次收窄至±10%，400次进一步收窄至±5%，推荐关键测试用例至少运行20-50次
1
Sources
50%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
UnverifiedWhen testing LLM citation behavior, the same prompt should be run at least 20–50 times, ideally expanded to more than 100 times.73% similarUnverified该测试样本规模相对有限（5个模型、13个代码库），结论有待更大规模验证73% similarUnverified实验采用8:2比例划分数据，20%作为内部测试集，训练集内部使用5折交叉验证，并移除重复记录以避免数据泄漏72% similarUnverifiedThe accuracy data in the test has an error margin of approximately ±8–10 percentage points at n=100 sample size (based on a 95% confidence-level binomial confidence interval).72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/530938API
curl https://kongchang.com/api/v1/knowledge/claims/530938MCP
get_claim(id=530938)