paper-with-me

홈 › Papers

PyHessian: Neural Networks Through the Lens of the Hessian

2019-12-16 · Zhewei Yao, Amir Gholami, Kurt Keutzer, Michael Mahoney

We present PYHESSIAN, a new scalable framework that enables fast computation of Hessian (i.e., second-order derivative) information for deep neural networks. PYHESSIAN enables fast computations of the top Hessian eigenvalues, the Hessian trace, and the full Hessian eigenvalue/spectral density, and it supports distributed-memory execution on cloud/supercomputer systems and is available as open source. This general framework can be used to analyze neural network models, including the topology of the loss landscape (i.e., curvature information) to gain insight into the behavior of different models/optimizers. To illustrate this, we analyze the effect of residual connections and Batch Normalization layers on the trainability of neural networks. One recent claim, based on simpler first-order analysis, is that residual connections and Batch Normalization make the loss landscape smoother, thus making it easier for Stochastic Gradient Descent to converge to a good solution. Our extensive analysis shows new finer-scale insights, demonstrating that, while conventional wisdom is sometimes validated, in other cases it is simply incorrect. In particular, we find that Batch Normalization does not necessarily make the loss landscape smoother, especially for shallower networks.

📄 PDF Abstract BibTeX arXiv:1912.07145

Code (3)

amirgholami/pyhessian 공식 구현 pytorch
cxtraa/ngd_with_slt pytorch
rmojgani/LPINNs pytorch

Methods 이 논문이 사용한 방법론

Batch Normalization 설명 없음

Similar Papers 제목 키워드 기반

Local properties of neural networks through the lens of layer-wise Hessians

2025-10-20 · Maxim Bolshim, Alexander Kugaevskikh arxiv

We introduce a methodology for analyzing neural networks through the lens of layer-wise Hessian matrices. The local Hessian of each functional block (layer) is defined as the matrix of second derivatives of a scalar func…

FLeNS: Federated Learning with Enhanced Nesterov-Newton Sketch

2024-09-23 · Sunny Gupta, Mohit Jindal, Pankhi Kashyap, Pranav Jeevan 외

Federated learning faces a critical challenge in balancing communication efficiency with rapid convergence, especially for second-order methods. While Newton-type algorithms achieve linear convergence in communication ro…

Dimensionality ReductionEdge-computingFederated LearningSecond-order methods

Revisiting Zeroth-Order Hessian Approximation: A Single-Step Policy Optimization Lens

2026-05-29 · Junbin Qiu, Zhaowei Hong, Renzhe Xu, Yao Shu arxiv

Accurate Zeroth-Order (ZO) Hessian estimation is a cornerstone of derivative-free methods, essential for tasks such as bilevel optimization, Bayesian inference, and uncertainty quantification. However, obtaining a comple…

Bilevel OptimizationBayesian Inference

Why Transformers Need Adam: A Hessian Perspective

2024-02-26 · Yushun Zhang, Congliang Chen, Tian Ding, Ziniu Li 외

SGD performs worse than Adam by a significant margin on Transformers, but the reason remains unclear. In this work, we provide an explanation through the lens of Hessian: (i) Transformers are "heterogeneous": the Hessian…

Hessian-based Analysis of Large Batch Training and Robustness to Adversaries

2018-02-22 · NeurIPS 2018 12 · Zhewei Yao, Amir Gholami, Qi Lei, Kurt Keutzer 외

Large batch size training of Neural Networks has been shown to incur accuracy loss when trained with the current methods. The exact underlying reasons for this are still not completely understood. Here, we study large ba…