AI Paper Atlas

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

Physics vs Distributions: Pareto Optimal Flow Matching with Physics Constraints

Giacomo Baldan, Qiang Liu, Alberto Guardone, Nils Thuerey

ICLR 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Flow matching and flow sampling · 29 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
train with PBFM (residual loss, ConFIG, unrolling, stochastic sampler) instead of plain FM-OTflow matching DiT · Darcy flowresidual MSE: 4.159 -> 0.838Improved
train with PBFM instead of plain FM-OTflow matching DiT · Darcy flowWasserstein distance x1e2: 0.059 -> 0.138Hurt
train with PBFM instead of plain FM-OTflow matching DiT · Kolmogorov flowresidual MSE x1e1: 2.314 -> 1.362Improved
train with PBFM instead of plain FM-OTflow matching DiT · Dynamic stallresidual MSE x1e6: 11.02 -> 0.339Improved
use stochastic sampler (t*=0.2) instead of deterministic sampler (ablation)PBFM · Darcy flowWasserstein distance x1e2: 1.470 -> 0.138Improved
increase noise resampling threshold t* from 0.2 to 1.0 (ablation)PBFM stochastic sampler · Darcy flowresidual vs WD: RE 0.838 -> 0.632; WD 0.138 -> 0.316Mixed
add ConFIG conflict-free gradients alone (no unrolling) (ablation)PBFM training · Darcy flowresidual MAEImproved
add unrolling steps during training (ablation)PBFM · Darcy flowresidual MAEImproved
increase sigma_min from 0 to 1e-4/1e-3/1e-2 (ablation)PBFM · Dynamic stallresidual and WD: RE 0.339 -> 0.473; WD 1.814 -> 3.246 (1e-3); MMSE 1.490 -> 1.200 (1e-2)Mixed
increase unrolling steps from 1 to 4 (ablation)PBFM training cost · Darcy flowtraining memory GB: 12.62 -> 44.15 GBHurt
use PBFM instead of PCFMinference · Dynamic stallinference time ms: 3906 -> 60.47Improved

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