paper-with-me

Papers

NysAct: A Scalable Preconditioned Gradient Descent using Nystrom Approximation

2025-06-10 · Hyunseok Seung, Jaewoo Lee, Hyunsuk Ko

Adaptive gradient methods are computationally efficient and converge quickly, but they often suffer from poor generalization. In contrast, second-order methods enhance convergence and generalization but typically incur high computational and memory costs. In this work, we introduce NysAct, a scalable first-order gradient preconditioning method that strikes a balance between state-of-the-art first-order and second-order optimization methods. NysAct leverages an eigenvalue-shifted Nystrom method to approximate the activation covariance matrix, which is used as a preconditioning matrix, significantly reducing time and memory complexities with minimal impact on test accuracy. Our experiments show that NysAct not only achieves improved test accuracy compared to both first-order and second-order methods but also demands considerably less computational resources than existing second-order methods. Code is available at https://github.com/hseung88/nysact.

📄 PDF Abstract BibTeX arXiv:2506.08360

Code (1)

hseung88/nysact 공식 구현 pytorch

Tasks

Second-order methods

Similar Papers 제목 키워드 기반

Dual Natural Gradient Descent for Scalable Training of Physics-Informed Neural Networks

2025-05-27 · Anas Jnini, Flavio Vella

Natural-gradient methods markedly accelerate the training of Physics-Informed Neural Networks (PINNs), yet their Gauss--Newton update must be solved in the parameter space, incurring a prohibitive $O(n^3)$ time complexit…

GPU

On the Nystrom Approximation for Preconditioning in Kernel Machines

2023-12-06 · Amirhesam Abedsoltan, Parthe Pandit, Luis Rademacher, Mikhail Belkin

Kernel methods are a popular class of nonlinear predictive models in machine learning. Scalable algorithms for learning kernel models need to be iterative in nature, but convergence can be slow due to poor conditioning. …

Recurrent neural network training with preconditioned stochastic gradient descent

2016-06-14 · Xi-Lin Li

This paper studies the performance of a recently proposed preconditioned stochastic gradient descent (PSGD) algorithm on recurrent neural network (RNN) training. PSGD adaptively estimates a preconditioner to accelerate g…

Handwritten Digit Recognition

Transformers learn to implement preconditioned gradient descent for in-context learning

2023-09-21 · NeurIPS 2023 11

Motivated by the striking ability of transformers for in-context learning, several works demonstrate that transformers can implement algorithms like gradient descent. By a careful construction of weights, these works sho…

Meta-Learning with a Geometry-Adaptive Preconditioner

2023-04-04 · CVPR 2023 1 · Suhyun Kang, Duhun Hwang, Moonjung Eo, Taesup Kim 외

Model-agnostic meta-learning (MAML) is one of the most successful meta-learning algorithms. It has a bi-level optimization structure where the outer-loop process learns a shared initialization and the inner-loop process …

Few-Shot Image ClassificationFew-Shot LearningMeta-Learning