Unverified50% confidenceSolutionExact time
Adversarial test cases should be written as deterministic binary pass/fail checkers rather than using fuzzy similarity scores.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe benchmark uses a deterministic checker for scoring—answers are either correct or incorrect—and does not use a large model as a judge.80% similarUnverified多种子重复实验中可使用配对t检验或Wilcoxon符号秩检验比较两组准确率分布，其中Wilcoxon检验因不依赖正态性假设更适合小样本78% similarUnverifiedThe tests use a combination of boolean pass/fail judgments and numerical scoring, ultimately aggregated into a single Overall score.76% similarUnverifiedMcNemar 检验要求两个模型在完全相同的测试集上评估，任何测试集的不一致都会使检验失效76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1000831API
curl https://kongchang.com/api/v1/knowledge/claims/1000831MCP
get_claim(id=1000831)