Unverified50% confidenceFactExact time
The rankings produced by pass@20 and all-20 are nearly opposite, revealing that strong capability does not imply high execution stability.
1
Sources
50%
Confidence
Long-term
Relevance
10/9/2026
First Seen
Sources
Related Entities
Related Claims
Unverified研究表明，即使能力相对较弱的模型，其pass@10或pass@20的表现往往能够接近强模型的pass@1水平71% similarUnverifiedIn highly competitive top-conference submissions, acceptance rates are often only 20%–25%.71% similarUnverified同一个精心设计的提示词在不同平台间直接复用时，输出质量可能出现20-40%的效果落差64% similarUnverified讨论者的共识是如果基座模型在pass@16上的成功率为0，RL几乎不可能凭空创造出新能力，但只要存在非零成功概率RL就能将其增强放大64% similarUnverifiedResearch shows that different prompt strategies on the same model can lead to differences in recall of more than 20 percentage points.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/991241API
curl https://kongchang.com/api/v1/knowledge/claims/991241MCP
get_claim(id=991241)