Input Space Mode Connectivity in Deep Neural Networks

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26620%2F26%3A0198121" target="_blank" >RIV/00216305:26620/26:0198121 - isvavai.cz</a>

Result on the web

<a href="https://openreview.net/pdf?id=3qeOy7HwUT" target="_blank" >https://openreview.net/pdf?id=3qeOy7HwUT</a>

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Input Space Mode Connectivity in Deep Neural Networks

Original language description

We extend the concept of loss landscape mode connectivity to the input space of deep neural networks. Mode connectivity was originally studied within parameter space, where it describes the existence of low-loss paths between different solutions (loss minimizers) obtained through gradient descent. We present theoretical and empirical evidence of its presence in the input space of deep networks, thereby highlighting the broader nature of the phenomenon. We observe that different input images with similar predictions are generally connected, and for trained models, the path tends to be simple, with only a small deviation from being a linear path. Our methodology utilizes real, interpolated, and synthetic inputs created using the input optimization technique for feature visualization. We conjecture that input space mode connectivity in high-dimensional spaces is a geometric effect that takes place even in untrained models and can be explained through percolation theory. We exploit mode connectivity to obtain new insights about adversarial examples and demonstrate its potential for adversarial detection. Additionally, we discuss applications for the interpretability of deep networks.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

<a href="/en/project/FW11020175" target="_blank" >FW11020175: Ecosystem for ensuring autonomous safety in Industry 4.0</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year

2025

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

ICLR 2025 The Thirteenth International Conference on Learning Representations

ISBN

9798331320850

ISSN

—

e-ISSN

—

Number of pages

25

Pages from-to

6394-6418

Publisher name

International Conference on Learning Representations, ICLR

Place of publication

Singapore

Event location

Singapore

Event date

Apr 24, 2025

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—