Tracking the Change of Knowledge Through Layers in Neural Networks

Xu Cheng, Lei Cheng, Zhaoran Peng, Quanshi Zhang

Keywords: Explainability, Forward Propagation, Neural Network, Representation Learning

TL;DR: This paper explains how a DNN gradually extracts new knowledge and forgets old knowledge through layers in forward propagation.

Abstract:

This paper aims to explain how a deep neural network (DNN) gradually extracts new knowledge and forgets noisy features
through layers in forward propagation.
Up to now, although how to define knowledge encoded by the DNN has not reached a consensus so far, previous studies have derived a series of mathematical evidences to take interactions as symbolic primitive inference patterns encoded by a DNN.
We extend the definition of interactions and, for the first time, extract interactions encoded by intermediate layers.
We quantify and track the newly emerged interactions and the forgotten interactions in each layer during the forward propagation, which shed new light on the learning behavior of DNNs.
The layer-wise change of interactions also reveals the change of the generalization capacity and instability of feature representations of a DNN.

Primary Area: neurosymbolic & hybrid AI systems (physics-informed, logic & formal reasoning, etc.)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 1270