stat.ML · 2026-10-02 · No. 131

Machine Learning, 2026-10-02.

3 new papers in stat.ML. Titles, authors, abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →

01 — The papers

3 entries
  1. 01

    Wasserstein Gradient Flows and Forward-Only Diffusion Are Not Enough for Multimodal Sampling

    Daniel McBride, Pratik Khandagale, Cristina Garcia-Cardona, Yen Ting Lin

    stat.ML · cs.LG · math-ph · math.DS · math.PR · math.SP

    There has been a proliferation of sampling algorithms based on Wasserstein gradient flows (WGF) and forward-only diffusion processes (FODP), often accompanied by theoretical guarantees of exponentially fast convergence to the target distribution. These guarantees are frequently interpreted as evidence that such methods can efficiently sample complex multimodal distributions, often supported by empirical results. In this work, we argue that...

    arxiv.org/abs/2610.02081 · PDF

  2. 02

    Error-Corrected Inference-Time Scaling for Imperfect Diffusion Models

    Zuokai Wen, Louis Grenioux, Weinan E, Jiequn Han

    stat.ML · cs.LG

    Inference-time scaling adapts pretrained diffusion models to new sampling tasks without additional training. Existing methods rely primarily on Monte Carlo sampling with more particles, yet are premised on the pretrained model being exact. In practice, data and training limitations make the model imperfect, and these methods inherit its error. More particles reduce Monte Carlo error but cannot remove the mismatch between the endpoint and the...

    arxiv.org/abs/2610.01933 · PDF

  3. 03

    The hidden advantage of mask resampling: a theory of masked autoencoders

    Jorge Medina Moreira, Lorenzo Bardone, Lenka Zdeborová

    stat.ML · cond-mat.dis-nn · cs.LG

    Why can masked prediction learn useful representations that unmasked reconstruction misses? We study this question in a high-dimensional model of a masked autoencoder (MAE) trained on data with shared latent structure and heterogeneous noise. We prove that masked linear reconstruction can recover the latent feature at linear sample complexity in regimes where unmasked linear reconstruction, equivalent to PCA, fails. The analysis also...

    arxiv.org/abs/2610.01578 · PDF

This edition is part of The Daily Abstract — stat.ML archive. Subscribe to receive these in your inbox each morning, automatically translated to Spanish, with reply-to-PDF: arxivdaily.ignorelist.com.

Colophon Set in Georgia, with system sans for interface chrome and a monospaced stack for code and paper identifiers. Sole accent: amber #D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.