Unverified50% confidenceFactExact time
Holm correction is a multiple-comparison correction method that controls the family-wise error rate: p values are sorted from smallest to largest, the k-th smallest p value is compared with α/(m−k+1), and rejections proceed stepwise.
1
Sources
50%
Confidence
Long-term
Relevance
11/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedHolm correction is a family-wise error rate (FWER) control method for multiple hypothesis testing that is less conservative and more statistically powerful than Bonferroni correction while still strictly controlling FWER.75% similarUnverifiedTop-p (nucleus sampling) limits the candidate token range by dynamically truncating probability mass, and has better adaptability than fixed-truncation Top-k66% similarUnverifiedThe paper attributes the inability of prior methods to distinguish different confidence targets to two problems: the reward signals that distinguish different p values are often lost during optimization, and the raw error penalty grows unboundedly as p approaches 1.65% similarUnverified对MoE模型的量化需要格外谨慎，因稀疏激活使专家参数数值分布集中，低比特量化易造成路由偏差导致能力退化64% similarUnverifiedBalanced accuracy equals the arithmetic mean of the recall rates of each category and is used to correct for multi-class imbalance issues.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1129341API
curl https://kongchang.com/api/v1/knowledge/claims/1129341MCP
get_claim(id=1129341)