AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Physics-Informed Distillation of Diffusion Models for PDE-Constrained Generation

Yi Zhang, Peng Wang, Difan Zou

ICML 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration · 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
add hard-constraint replacement of masked entries at inference (ablation)PIDDM · Darcy forwardPDE error (x1e-4): 0.354 -> 0.145Improved
add hard-constraint replacement at inference (ablation)PIDDM · Darcy forwardMSE (x1e-1): 0.705 -> 0.316Improved
add two rounds of reflowing (ablation)PIDDM · Darcy forwardMSE (x1e-1): 0.316 -> 0.127Improved
add consistency model distillation (ablation)PIDDM · Darcy forwardPDE error (x1e-4): 0.145 -> 0.083Improved
add distribution matching distillation (ablation)PIDDM · Darcy forwardMSE (x1e-1): 0.316 -> 0.255Improved
use VP diffusion process instead of linear (ablation)PIDDM · Darcy forwardMSE (x1e-1): 0.316 -> 0.398Hurt
use sub-VP diffusion process instead of linear (ablation)PIDDM · Darcy forwardMSE (x1e-1): 0.316 -> 0.372Hurt
add optional refinement of the latent noise (PIDDM-ref) (ablation)PIDDM-1 · Darcy generationPDE error (x1e-4): 0.226 -> 0.148Improved
add refinement of the latent noise (PIDDM-ref) (ablation)PIDDM-1 · Darcy generationFPD: 0.754 -> 0.385Improved
increase teacher sampling steps N_s (ablation)PIDDM student · DarcyMMSE, SMSE, PDE errorImproved
vary PDE loss weight in distillation and inference (ablation)PIDDM · Darcyall metricsNo clear effect
enforce PDE on final sample via distillation instead of posterior mean (PIDM)PIDDM-1 · Darcy generationPDE error (x1e-4): 1.236 -> 0.226Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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