AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Variational Flow Maps: Make Some Noise for One-Step Conditional Generation

Abbas Mammadov, So Takao, Bohan Chen, Ricardo Baptista, Morteza Mardani, Yee Whye Teh, Julius Berner

ICML 2026 · Image generation › Fast diffusion sampling · approach: Flow matching distillation · 0 citations (OpenAlex)

Read the paper

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
jointly train flow map with the noise adapter instead of freezing theta (ablation)VFM · ImageNet 256 box inpaintingFID: 136.12 -> 33.34Improved
jointly train flow map with the noise adapter instead of freezing theta (ablation)VFM · ImageNet 256 Gaussian deblurringFID: 172.74 -> 51.05Improved
add the mean flow loss to constrain theta (vs unconstrained-theta) (ablation)VFM · 2D checkerboard posterioroff-manifold samplesImproved
add the KL term on the adapter (ablation)VFM · 2D checkerboardtraining stabilityImproved
use an EMA of theta in the observation loss (ablation)VFM · 2D checkerboardsupport accuracyImproved
increase tau to 1 or above (from tau at or below sigma) (ablation)VFM · 2D checkerboardCRPS and posterior MMDImproved
set alpha to 1 versus below 1 (ablation)VFM · 2D checkerboardposterior vs prior MMDMixed
use alpha < 1 (0.5) for unconditional generation (ablation)VFM · ImageNet 256 unconditionalFID: FID below 10 with 2-stepImproved
average 10 VFM posterior samples instead of a single sampleVFM · ImageNet 256 Gaussian deblurringPSNR: 21.74 -> 23.92Improved
average 10 VFM posterior samples instead of a single sampleVFM · ImageNet 256 box inpaintingPSNR: 21.98 -> 22.71Improved
use VFM instead of guidance baselinesLatent DPS · ImageNet 256 box inpaintingFID: 62.89 -> 33.34Improved
use VFM single sample instead of Latent DAPSLatent DAPS · ImageNet 256 box inpaintingPSNR: 23.98 -> 21.98Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 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