AI Paper Atlas

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

Minibatch Optimal Transport and Perplexity Bound Estimation in Discrete Flow Matching

Etrit Haxholli, Yeti Z. Gurbuz, Oğul Can, Eli Waxman

ICML 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Flow matching and flow sampling · 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 minibatch OT (Sinkhorn) to BoW discrete flow (ablation)DFM-B · OWT generative perplexity, 1024 stepsgenerative perplexity: 185.12 -> 178.53Improved
add minibatch OT (Sinkhorn) to multimask flow (ablation)DFM-MMF · OWT generative perplexity, 1024 stepsgenerative perplexity: 143.48 -> 141.92Improved
replace Sinkhorn with exact OT solver (ablation)DFM-MMF · OWT generative perplexity, 1024 stepsgenerative perplexity: 141.92 -> 141.46No clear effect
compute OT couplings with EMA embeddings (OT-EMA) on top of exact OT (ablation)DFM-MMF-Exact · OWT generative perplexity, 1024 stepsgenerative perplexity: 141.46 -> 132.55Improved
compute OT couplings with EMA embeddings (OT-EMA) on top of exact OT (ablation)DFM-B-Exact · OWT generative perplexity, 8 stepsgenerative perplexity: 335.14 -> 302.76Improved
use multimask flow instead of single-mask source (ablation)masked DFM · OWT generative perplexity, 1024 stepsgenerative perplexity: 147.07 -> 143.48Improved
use BoW source instead of multimask source (ablation)DFM with OT · OWT generative perplexity, few vs many stepsgenerative perplexity: 8 steps: 302.76 vs 480.61; 1024 steps: 159.87 vs 132.55Mixed
add minibatch OT to Morse Shakespeare flowGPT-2 DFM, BoW source · Shakespeare Morse-code, L=128jumps: 85.47 -> 74.87Improved
increase Sinkhorn entropy regularization epsilon (ablation)minibatch OT · Shakespeare Morse-code, L=128jumps: 74.87 -> 82.86Hurt
reduce number of mask tokens to 256 (ablation)DFM-MMF with OT · OWTgenerative perplexityHurt
increase OT batch size to 4096 (ablation)minibatch OT · OWT, 32 and 64 stepsgenerative perplexityImproved
train with the (1-t)-weighted bound (DFM-N) rather than MD4-style loss (DFM-S) (ablation)masked DFM · OWT generative perplexity, 8 stepsgenerative perplexity: 587.80 -> 556.73Improved

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