Differential Geometric View of Information Flow in Neural Nets

Suhas Sreehari, Pradeep Ramuhalli, Frank Liu

Research output: Contribution to journalConference articlepeer-review

Abstract

In this paper, we explore a space-time geometric view of signal representation in machine learning models. The question we are interested in is if we can identify what is causing signal representation errors – training data inadequacies, model insufficiencies, or both. Loosely expressed, this problem is stylistically similar to blind deconvolution problems. However, studies of space-time geometries might be able to partially solve this problem by considering the curvature produced by mass in (Anti-)de Sitter space. We study the effectiveness of our approach on the MNIST dataset.

Original languageEnglish
Article number141
JournalIS and T International Symposium on Electronic Imaging Science and Technology
Volume37
Issue number14
DOIs
StatePublished - 2025
EventIS and T International Symposium on Electronic Imaging 2025: 23rd Computational Imaging, COIMNG 2025 - Burlingame, United States
Duration: Feb 2 2025Feb 6 2025

Fingerprint

Dive into the research topics of 'Differential Geometric View of Information Flow in Neural Nets'. Together they form a unique fingerprint.

Cite this