Per-example gradients: a new frontier for understanding and improving optimizers

• URL: http://arxiv.org/abs/2510.00236v1
• Date: Tue, 30 Sep 2025 20:00:41 GMT
• Title: Per-example gradients: a new frontier for understanding and improving optimizers
• Authors: Vincent Roulet, Atish Agarwala,
• Abstract summary: We show that gradient statistics can be implemented through a surgery of the automatic differentiation graph.<n>We also revise our understanding of two nonlinear operations in optimization through the lens of per-example gradient transformations.<n>Overall we demonstrate that per-example gradient information enables new analyses and possibilities for algorithm design.
• Score: 10.653229860484464
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Training algorithms in deep learning usually treat a mini-batch of samples as a single object; they average gradients over the mini-batch, and then process the average in various ways. Computing other statistics beyond the average may have been seen as prohibitively resource intensive in automatic differentiation (AD) frameworks. We show that this is not the case. Generally, gradient statistics can be implemented through a surgery of the AD graph, which, in some cases, incur almost no computational and memory overheads compared to the mini-batch gradient computation. Additionally, we show that in certain classes of models, including transformers, JAX's vectorization transformation offers a viable implementation for prototyping and experimentation. We then revise our understanding of two nonlinear operations in optimization through the lens of per-example gradient transformations. We first study signSGD and show that the optimal placement of the sign operation in the gradient processing chain is crucial to success and can be predicted with a simple signal-to-noise ratio argument. Next we study per-example variations of the Adam preconditioner, and show that optimization is best served when the preconditioner is dominated by the mean rather than the variance of the gradient distribution - in contrast to conventional wisdom. Overall we demonstrate that per-example gradient information enables new analyses and possibilities for algorithm design.

Related papers

• GradMetaNet: An Equivariant Architecture for Learning on Gradients [18.350495600116712]
  We introduce GradMetaNet, a novel architecture for learning on gradients.<n>We prove results for GradMetaNet, and show that previous approaches cannot approximate natural gradient-based functions.<n>We then demonstrate GradMetaNet's effectiveness on a diverse set of gradient-based tasks.
  arXiv  Detail & Related papers  (2025-07-02T12:22:39Z)
• Revisiting the Initial Steps in Adaptive Gradient Descent Optimization [6.468625143772815]
  Adaptive gradient optimization methods, such as Adam, are prevalent in training deep neural networks across diverse machine learning tasks.<n>These methods often suffer from suboptimal generalization compared to descent gradient (SGD) and exhibit instability.<n>We introduce simple yet effective solutions: initializing the second-order moment estimation with non-zero values.
  arXiv  Detail & Related papers  (2024-12-03T04:28:14Z)
• Efficient Sharpness-Aware Minimization for Molecular Graph Transformer Models [42.59948316941217]
  Sharpness-aware minimization (SAM) has received increasing attention in computer vision since it can effectively eliminate the sharp local minima from the training trajectory and generalization degradation. We propose a new algorithm named GraphSAM, which reduces the training cost of SAM and improves the generalization performance of graph transformer models.
  arXiv  Detail & Related papers  (2024-06-19T01:03:23Z)
• Scaling Forward Gradient With Local Losses [117.22685584919756]
  Forward learning is a biologically plausible alternative to backprop for learning deep neural networks. We show that it is possible to substantially reduce the variance of the forward gradient by applying perturbations to activations rather than weights. Our approach matches backprop on MNIST and CIFAR-10 and significantly outperforms previously proposed backprop-free algorithms on ImageNet.
  arXiv  Detail & Related papers  (2022-10-07T03:52:27Z)
• Adapting Stepsizes by Momentumized Gradients Improves Optimization and Generalization [89.66571637204012]
  textscAdaMomentum on vision, and achieves state-the-art results consistently on other tasks including language processing. textscAdaMomentum on vision, and achieves state-the-art results consistently on other tasks including language processing. textscAdaMomentum on vision, and achieves state-the-art results consistently on other tasks including language processing.
  arXiv  Detail & Related papers  (2021-06-22T03:13:23Z)
• Zeroth-Order Hybrid Gradient Descent: Towards A Principled Black-Box Optimization Framework [100.36569795440889]
  This work is on the iteration of zero-th-order (ZO) optimization which does not require first-order information. We show that with a graceful design in coordinate importance sampling, the proposed ZO optimization method is efficient both in terms of complexity as well as as function query cost.
  arXiv  Detail & Related papers  (2020-12-21T17:29:58Z)
• Channel-Directed Gradients for Optimization of Convolutional Neural Networks [50.34913837546743]
  We introduce optimization methods for convolutional neural networks that can be used to improve existing gradient-based optimization in terms of generalization error. We show that defining the gradients along the output channel direction leads to a performance boost, while other directions can be detrimental.
  arXiv  Detail & Related papers  (2020-08-25T00:44:09Z)
• Randomized Automatic Differentiation [22.95414996614006]
  We develop a general framework and approach for randomized automatic differentiation (RAD) RAD can allow unbiased estimates to be computed with reduced memory in return for variance. We show that RAD converges in fewer iterations than using a small batch size for feedforward networks, and in a similar number for recurrent networks.
  arXiv  Detail & Related papers  (2020-07-20T19:03:44Z)
• Variance Reduction with Sparse Gradients [82.41780420431205]
  Variance reduction methods such as SVRG and SpiderBoost use a mixture of large and small batch gradients. We introduce a new sparsity operator: The random-top-k operator. Our algorithm consistently outperforms SpiderBoost on various tasks including image classification, natural language processing, and sparse matrix factorization.
  arXiv  Detail & Related papers  (2020-01-27T08:23:58Z)
• Towards Better Understanding of Adaptive Gradient Algorithms in Generative Adversarial Nets [71.05306664267832]
  Adaptive algorithms perform gradient updates using the history of gradients and are ubiquitous in training deep neural networks. In this paper we analyze a variant of OptimisticOA algorithm for nonconcave minmax problems. Our experiments show that adaptive GAN non-adaptive gradient algorithms can be observed empirically.
  arXiv  Detail & Related papers  (2019-12-26T22:10:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.