Saved in:
Bibliographic Details
Main Authors: Sbailò, Luigi, Ghiringhelli, Luca
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2310.08224
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912223685771264
author Sbailò, Luigi
Ghiringhelli, Luca
author_facet Sbailò, Luigi
Ghiringhelli, Luca
contents The configuration of latent representations plays a critical role in determining the performance of deep neural network classifiers. In particular, the emergence of well-separated class embeddings in the latent space has been shown to improve both generalization and robustness. In this paper, we propose a method to induce the collapse of latent representations belonging to the same class into a single point, which enhances class separability in the latent space while enforcing Lipschitz continuity in the network. We demonstrate that this phenomenon, which we call \textit{latent point collapse}, is achieved by adding a strong $L_2$ penalty on the penultimate-layer representations and is the result of a push-pull tension developed with the cross-entropy loss function. In addition, we show the practical utility of applying this compressing loss term to the latent representations of a low-dimensional linear penultimate layer. The proposed approach is straightforward to implement and yields substantial improvements in discriminative feature embeddings, along with remarkable gains in robustness to input perturbations.
format Preprint
id arxiv_https___arxiv_org_abs_2310_08224
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Latent Point Collapse on a Low Dimensional Embedding in Deep Neural Network Classifiers
Sbailò, Luigi
Ghiringhelli, Luca
Machine Learning
The configuration of latent representations plays a critical role in determining the performance of deep neural network classifiers. In particular, the emergence of well-separated class embeddings in the latent space has been shown to improve both generalization and robustness. In this paper, we propose a method to induce the collapse of latent representations belonging to the same class into a single point, which enhances class separability in the latent space while enforcing Lipschitz continuity in the network. We demonstrate that this phenomenon, which we call \textit{latent point collapse}, is achieved by adding a strong $L_2$ penalty on the penultimate-layer representations and is the result of a push-pull tension developed with the cross-entropy loss function. In addition, we show the practical utility of applying this compressing loss term to the latent representations of a low-dimensional linear penultimate layer. The proposed approach is straightforward to implement and yields substantial improvements in discriminative feature embeddings, along with remarkable gains in robustness to input perturbations.
title Latent Point Collapse on a Low Dimensional Embedding in Deep Neural Network Classifiers
topic Machine Learning
url https://arxiv.org/abs/2310.08224