Understanding the role of individual units in a deep neural network
Name
30071.full.pdf
Description
Published version
Size
6.51 MB
Format
Adobe PDF
Checksum (MD5)
400bd8a9dc42c9acd7188833c1b9f4aa
Author(s)
Bau, David
Zhu, Jun-Yan
Strobelt, Hendrik
Lapedriza Garcia, Agata
Zhou, Bolei
Torralba, Antonio
Date Issued
September 2020
Journal
Proceedings of the National Academy of Sciences
Publisher
Proceedings of the National Academy of Sciences
Citation
Bau, David et al. "Understanding the role of individual units in a deep neural network." Proceedings of the National Academy of Sciences 117, 48 (September 2020): 30071-30078 © 2020 National Academy of Sciences
Version
Final published version
Abstract
Deep neural networks excel at finding hierarchical representations that solve complex tasks over large datasets. How can we humans understand these learned representations? In this work, we present network dissection, an analytic framework to systematically identify the semantics of individual hidden units within image classification and image generation networks. First, we analyze a convolutional neural network (CNN) trained on scene classification and discover units that match a diverse set of object concepts. We find evidence that the network has learned many object classes that play crucial roles in classifying scene classes. Second, we use a similar analytic method to analyze a generative adversarial network (GAN) model trained to generate scenes. By analyzing changes made when small sets of units are activated or deactivated, we find that objects can be added and removed from the output scenes while adapting to the context. Finally, we apply our analytic framework to understanding adversarial attacks and to semantic image editing.
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Massachusetts Institute of Technology. Media Laboratory
MIT-IBM Watson AI Lab
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1073/pnas.1907375117
