AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Exploring and Exploiting Stability in Latent Flow Matching

Rania Briq, Michael Kamp, Ohad Fried, Sarel Cohen, Stefan Kesselheim

ICML 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Flow matching generative modeling · 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
prune training data with balanced clustering (Cb) at pr=0.5DiT-S/2 LFM · CelebA-HQFID: 24.24 -> 22.80Improved
select samples with highest loss (L) or lowest gradient (G-1) instead of the inverse (ablation)LFM pruning · CelebA-HQ pr=0.5FID: 24.24 -> 33.92 (L), 29.75 (G-1)Hurt
use coreset selection instead of balanced clustering (Cb) (ablation)LFM pruning · CelebA-HQ pr=0.5FID: 22.80 -> 26.94 (balanced coreset), 26.25 (global coreset)Hurt
select cluster samples furthest from center instead of nearest (ablation)Cb pruning · CelebA-HQ pr=0.5FID: 22.80 -> 26.77Hurt
use balanced instead of proportional clustering (ablation)clustering pruning · CelebA-HQ pr=0.5FID: 25.19 -> 22.80Improved
random pruning compared with unpruned (ablation)LFM training · CelebA-HQ pr=0.5FID: 24.24 -> 25.25±0.38Hurt
scale model from DiT-S/2 to DiT-XL/2 (20x weights) (ablation)LFM · CelebA-HQFID: 24.24 -> 22.87Improved
switch backbone from DiT to U-Net (ablation)LFM · CelebA-HQArcFace similarity s: s = 0.55Hurt
replace flow matching with score-based diffusion (ablation)LFM · CelebA-HQoutput stabilityHurt
change the VAE (training seed) or apply invertible sign flip to latent space (ablation)LFM · CelebA-HQArcFace similarity s: s = 0.32 (VAE swap), 0.48 (flip first feature), 0.32 (flip all)Hurt
add coarse-to-fine two-model inference (t0 = 0.7)DiT-XL/2 LFM · CelebA-HQ, H100, batch 128, 256x256inference cost (ms/img): 93.95 -> 43.53 ms/img (≈2.15x faster)Improved
train Coarse on a single perceived gender (stability violated) instead of balanced pruned data (ablation)coarse-to-fine model · CelebA-HQFID: FID = 44.92Hurt

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