Unverified50% confidenceFactExact time
The 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)).
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedUnder the assumption that the true mapping is uniformly random, the paper proves a lower bound on the error rate of E ≥ (M/N)δ*(B/M) + (1-M/N)(1-1/K).77% 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)73% similarUnverifiedMcNemar 检验统计量为 χ² = (|b - c| - 1)² / (b + c)，当样本量足够时近似服从自由度为1的卡方分布，其中-1为Yates连续性校正项70% similarUnverifiedThe verification process in traditional speculative decoding accepts draft-proposed tokens with probability min(1, p(x)/q(x)), and upon rejection resamples from the corrected distribution norm(max(0, p(x)-q(x))), which can be proven to produce a token distribution identical to directly sampling from the target model (lossless).69% similarUnverifiedThe error introduced by uniform quantization approximately follows a uniform distribution, with a variance of Δ²/12 (where Δ is the quantization step size).69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/976005API
curl https://kongchang.com/api/v1/knowledge/claims/976005MCP
get_claim(id=976005)