CORTEXA
← Browse
arxivstat.MLcs.LGmath.PR2026-07-18

Dropout and Random Gradient Masking Are Asymptotically Equivalent in Large ResNets

Javier Maass, Lénaïc Chizat

Dropout and Random Gradient Masking (RaM) are two training techniques used to improve performance in deep learning. Both techniques inject randomness into the training dynamics, but in significantly different ways: dropout applies random masks to the activations in the forward pass, whereas RaM leaves the forward pass unchanged and instead masks the gradients. In particular, the noise induced by RaM in the parameter updates is unbiased, so standard explanations for the effectiveness of dropout, such as the penalization effect or the prevention of co-adaptation between neurons, do not apply to RaM. In this work, we show that the difference between the two methods disappears for ResNets in the large depth and width asymptotics: in the complete feature learning regime, they both converge to the same large-scale limiting dynamics. This asymptotic equivalence holds for several variants of dropout and RaM, including layerwise dropout as used in stochastic-depth ResNets, albeit at slower quantitative rates. In fact, we also show that several of these variants collapse to the same limit asymptotically.

View free PDFSource page

Related papers

arxivmath.OCcs.LGmath.PRstat.ML2026-06-30

Homogenization of $\ell_2$-Adversarial Training in High-Dimensions: Exact Dynamics under Stochastic Gradient Descent

Fabrizzio Sabelli

We develop a framework for analyzing the learning dynamics of $\ell_2$-adversarial training of single-index models on Gaussian mixtures in the high-dimensional limit under streaming stochastic gradient descent (SGD). We derive deterministic equivalents for a broad class of statis…

View free PDFSource page
arxivcs.LGcs.DSmath.NAmath.PRstat.ML2026-06-26

VGB for Masked Diffusion Model: Efficient Test-time Scaling for Reward Satisfaction and Sample Editing

Kijung Jeon, Thuy-Duong Vuong, Molei Tao

Inference-time scaling is a promising paradigm to improve generative models, especially when outputs must satisfy structural constraints or optimize downstream rewards. We consider Masked Diffusion Model (MDM) and introduce MDM-VGB, a discrete diffusion sampler that augments unma…

View free PDFSource page
arxivstat.MLcs.LGmath.PR2026-06-27

Variance Reduction for Stochastic Gradient Generalized Non-reversible Langevin Monte Carlo Algorithms

Bingye Ni, Xiaoyu Wang, Yingli Wang, Lingjiong Zhu

We study the leading-order fluctuation of stochastic gradient Euler-Maruyama estimators for generalized non-reversible Langevin dynamics. Under structural assumptions tailored to the small-stepsize central limit theorem and under an unbiased stochastic gradient oracle, we prove t…

View free PDFSource page
arxivstat.MLcs.LGmath.PR2026-07-07

Fast determinantal sampling on general spaces and diffusion geometry

Hoang-Son Tran, Pranav Gupta, Subhroshekhar Ghosh

Determinantal point processes have recently emerged as a kernel-based alternative to standard independent sampling for constructing efficient minibatches, coresets, and other compact representations of large-scale datasets. In particular, sampling mechanisms based on DPPs are bel…

View free PDFSource page
arxivstat.COcs.LGmath.PRstat.ML2026-07-16

Delocalization of bias in unadjusted Hamiltonian Monte Carlo and underdamped Langevin

Yifan Chen, Xiaoou Cheng, Jonathan Niles-Weed, Jonathan Weare

Unadjusted samplers such as unadjusted Hamiltonian Monte Carlo and underdamped Langevin are well-known to be biased. Metropolis--Hastings adjustment has been conventionally incorporated into Hamiltonian Monte Carlo to eliminate the bias. However, this adjustment can significantly…

View free PDFSource page