Unverified50% confidenceFactExact time
The paper formalizes the blind spot via the A-equivalence class [P]_A: all distributions that have the same pushforward distribution as P under A form an equivalence class.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedChecking whether a benchmark is saturated requires looking at the overall distribution spread, not just the top score; if first and tenth place differ by only a hair, the leaderboard is saturated.64% similarUnverified二项分布的方差为Np(1-p)，在p=0.5时达到最大值63% similarUnverifiedLLM每次调用可建模为伯努利试验，N次调用通过次数服从二项分布B(N, p)，通过率95%置信区间约为p̂ ± 1.96×√(p̂(1-p̂)/N)61% similarUnverifiedThe paper proves that the learned generative distribution satisfies a finite-sample error bound in Wasserstein-p distance: W_p ≲ n^(-1/d) + n^(-1/(2p))(log(1/ξ))^(1/(2p)).61% similarUnverified当 b+c < 25 时应改用精确的二项分布检验，将 b 视为参数 p=0.5 的二项分布下的观测值61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1068375API
curl https://kongchang.com/api/v1/knowledge/claims/1068375MCP
get_claim(id=1068375)