AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures

Chenyang Wang, Weizhong Wang, Yinuo Ren, Jose Blanchet, Yiping Lu

ICML 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Hamiltonian Monte Carlo · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use pathwise Girsanov weights (URGE) instead of FK-Steering reward-difference weightsSMC guidance on GMM · 30-d GMMCov Frobenius error: 198.20 -> 181.31Improved
use derivative-free URGE instead of AFDPS+VCGGMM guidance · 30-d GMMMMD: 0.08 -> 0.06Improved
add resampling on top of pure guidance (AFDPS)Pure guidance · 30-d GMMMMD: 0.17 -> 0.10Improved
add variance-controlling guidance (VCG) to AFDPSAFDPS · 30-d GMMMMD: 0.10 -> 0.08Improved
increase number of particles for URGE (ablation)URGE · 30-d GMMaccuracy and runtimeImproved
use URGE instead of AFDPS variants on inverse problemsImageNet-256 inverse problems · ImageNet-256 inverse problemsPSNR/LPIPS: 22.38 vs 22.43 (Gaussian deblurring PSNR)No clear effect
use URGE instead of AFDPS-ODE on motion deblurringImageNet-256 motion deblurring · ImageNet-256 motion deblurringPSNR: 21.46 -> 18.37Hurt
use URGE instead of FK-Corrector on inverse problemsImageNet-256 Gaussian deblurring · ImageNet-256 Gaussian deblurringPSNR: 18.36 -> 22.38Improved
use URGE (N=4) instead of FK-Steering (N=4)SD v1.5 text-to-image · 50 promptsImageReward: 0.8397 -> 0.9955Improved
use URGE (N=4) instead of single sample N=1SD v1.5 text-to-image · 50 promptsGenEval: 0.6400 -> 0.7800Improved
add gradient guidance to FK-Steering (ablation)FK-Steering N=4 · 50 promptsImageReward: 0.8397 -> 0.7906Hurt
add gradient guidance to the N=1 sampler (ablation)SD v1.5 N=1 · 50 promptsImageReward: 0.2144 -> 0.2074No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy