technique
importance sampling
techniqueactive
importance-sampling-8e4d8272·1 events·first seen 26d agoAliases: importance sampling
Co-occurring entities
More like this (12)
Feature Importance MethodsBest-of-N SamplingTop-p samplingMin-p samplingsparse frame samplingimportance-weighted supervised fine-tuningNeural Super Samplingfinite-sample posterior sampling frameworkForce-Informed Re-Sampling Trainingdefer-to-resampledisagreement-focused samplingdiffusion posterior sampling
Recent events (1)
CARV: Compute-Aware Variance Reduction for Diffusion Teacher Gradient Estimation
CARV is a hierarchical Monte Carlo estimation framework that reduces gradient variance when using frozen pretrained diffusion models as teachers in downstream pipelines such as text-to-3D distillation and data attribution. The approach amortizes expensive upstream computation (rendering, simulation, encoding) over cheap diffusion-noise resamples, augmented by timestep importance sampling and stratified-inverse-CDF construction. In text-to-3D experiments, CARV delivers 2–3× effective compute multipliers; in single-step distillation, it cuts gradient variance by an order of magnitude but does not improve FID, revealing that MC variance is not the bottleneck in that regime.