跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-13AI 評分37

擴大流匹配中的線上最優傳輸小批次規模

Scaling Online Optimal Transport Minibatches for Flow Matching

AI 導讀

研究提出可擴展的線上熵正則最優傳輸近似,讓流匹配訓練能使用更大的來源—目標耦合小批次。理論證明,無正則化及熵正則最優傳輸均較獨立配對嚴格降低完整隨機梯度方差與積分誤差,且小批次增大時改善更明顯。單細胞基因插值流及圖像流實驗顯示,較大的線上耦合可提升訓練穩定性及最終生成表現。

正文

Published on Oct 13

Authors:

Abstract

A central challenge in scaling flow models is maintaining stability in training and integration. In flow matching, this stability depends critically on how source samples are paired with data. Independent random pairing introduces substantial stochasticity, motivating minibatch optimal transport (OT) to construct online source-target couplings that reduce this randomness (Tong et al., 2024). However, as applications extend to larger and higher-dimensional datasets, small minibatches can limit these benefits, while scaling the coupling to substantially larger minibatches at every iteration is computationally prohibitive. Moreover, how OT coupling affects the full stochastic gradient and integration remains theoretically unresolved. In this work, we develop a scalable online entropic OT approximation that enables substantially larger source-target couplings while preserving the benefits of OT. We close the theoretical gap for both unregularized and entropic OT by proving strict reductions in full stochastic-gradient variance and integration error relative to independent pairing, and further characterize how these improvements strengthen as the online coupling minibatch grows. Experiments on single-cell gene interpolation-flow and image-flow training validate these predictions, showing that larger online couplings consistently improve training stability and final generation performance.

View arXiv page View PDF Add to collection

Get this paper in your agent:

hf papers read 2610.18838

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.18838 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.18838 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.18838 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co