paper-with-me

홈 › Papers

How Gradient Descent Separates Data with Neural Collapse: A Layer-Peeled Perspective

2021-05-21 · NeurIPS 2021 12 · Wenlong Ji, Yiping Lu, Yiliang Zhang, Zhun Deng, Weijie J Su

In this paper, we derive a landscape analysis to the surrogate model to study the inductive bias of the neural features and parameters from neural networks with cross-entropy. We show that once the training cross-entropy loss decreases below a certain threshold, the features and classifiers in the last layer of the neural network will converge to a certain geometry structure, which is known as neural collapse\citep{papyan2020prevalence,fang2021layer}, \emph{i.e.} cross-example within-class variability of last-layer feature collapses to zero and the class-means converge to a Simplex Equiangular Tight Frame (ETF). We illustrate that the cross-entropy loss enjoys a benign global landscape where all the critical points are strict saddles whose Hessian exhibit negative curvature directions except the only global minimizers which exhibit neural collapse phenomenon.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

Inductive Bias

Similar Papers 제목 키워드 기반

Imitate Optimal Policy: Prevail and Induce Action Collapse in Policy Gradient

2025-09-02 · Zhongzhu Zhou, Yibo Yang, Ziyan Chen, Fengxiang Bie 외 arxiv

Policy gradient (PG) methods in reinforcement learning frequently utilize deep neural networks (DNNs) to learn a shared backbone of feature representations used to compute likelihoods in an action selection layer. Numero…

Reinforcement LearningOpenAI Gym

Deep Neural Collapse Is Provably Optimal for the Deep Unconstrained Features Model

2023-05-22 · NeurIPS 2023 11

Neural collapse (NC) refers to the surprising structure of the last layer of deep neural networks in the terminal phase of gradient descent training. Recently, an increasing amount of experimental evidence has pointed to…

Binary Classification

Neural Collapse versus Low-rank Bias: Is Deep Neural Collapse Really Optimal?

2024-05-23 · Peter Súkeník, Marco Mondelli, Christoph Lampert

Deep neural networks (DNNs) exhibit a surprising structure in their final layer known as neural collapse (NC), and a growing body of works has currently investigated the propagation of neural collapse to earlier layers o…

Binary ClassificationMulti-class Classification

Wide Neural Networks Trained with Weight Decay Provably Exhibit Neural Collapse

2024-10-07 · Arthur Jacot, Peter Súkeník, Zihan Wang, Marco Mondelli

Deep neural networks (DNNs) at convergence consistently represent the training data in the last layer via a highly symmetric geometric structure referred to as neural collapse. This empirical evidence has spurred a line …

Deriving Decoder-Free Sparse Autoencoders from First Principles

2026-01-10 · Alan Oursland arxiv

Gradient descent on log-sum-exp (LSE) objectives performs implicit expectation--maximization (EM): the gradient with respect to each component output equals its responsibility. The same theory predicts collapse without v…