Unverified50% confidenceFactExact time
Gaussian output distribution is the most fundamental form of predictive uncertainty quantification (UQ), where the model outputs distribution parameters (mean and variance) rather than a single scalar.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe residual distribution is defined as follows: for tokens where the target model probability p(x) is greater than the draft model probability q(x), the residual probability is proportional to max(0, p(x)-q(x)); this mechanism ensures that the marginal distribution of the sampled results strictly equals the original distribution of the target model.73% similarUnverifiedA prompt is essentially the 'condition' for the model; the more precise the condition, the more concentrated the model's probability distribution becomes on the user's desired output, rather than being spread across all directions.66% similarUnverified生成式模型建模联合概率 P(x,y) 或数据分布 P(x)，而判别式模型直接建模条件概率 P(y|x)；生成模型在小样本时更具优势，判别模型在大数据量时通常表现更好66% similarVerifiedThe core principle of large models is probabilistic text continuation, and the output is probabilistic sampling rather than deterministic output.66% similarUnverifiedProbability calibration refers to a post-processing technique that converts model output prediction scores into true probability estimates, with commonly used methods including Platt Scaling and Isotonic Regression.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/901106API
curl https://kongchang.com/api/v1/knowledge/claims/901106MCP
get_claim(id=901106)