Unverified50% confidenceFactExact time
SMD introduces a residual surrogate called Pivot, whose role is to separate local corrections from the student's interval-mean prediction: the student is responsible for predicting the average velocity over the entire interval, and Pivot is responsible for modeling the residual fluctuations within the interval.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLIME explains individual predictions by generating a perturbed dataset around the input sample to train a local linear approximation model, while SHAP provides feature attribution based on game-theoretic Shapley values; both are model-agnostic techniques.63% similarUnverifiedFUDGE trains a discriminator that predicts whether a future sequence will satisfy a given attribute, and re-weights candidate tokens at each step.63% similarUnverified知识蒸馏的中间层对齐策略（如PKD、TinyBERT）通过强制学生模型的隐藏表示模仿教师模型的分布间接约束嵌入几何结构61% similarUnverifiedOn-policy distillation (OPD) has the student model generate its own trajectories, and then the teacher model provides supervision signals on those trajectories.61% similarUnverifiedThe paper corresponding to SMD is titled "Simplifying and Improving On-Policy Distillation for Few-Step Diffusion", and its core contributions can be summarized as the joint design of the Pivot residual surrogate and the repulsion mechanism, and the cross-modality, cross-scale scalability that this brings.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1024029API
curl https://kongchang.com/api/v1/knowledge/claims/1024029MCP
get_claim(id=1024029)