ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 2106.12447
39
31

How Well do Feature Visualizations Support Causal Understanding of CNN Activations?

23 June 2021
Roland S. Zimmermann
Judy Borowski
Robert Geirhos
Matthias Bethge
Thomas S. A. Wallis
Wieland Brendel
    FAtt
ArXivPDFHTML
Abstract

A precise understanding of why units in an artificial network respond to certain stimuli would constitute a big step towards explainable artificial intelligence. One widely used approach towards this goal is to visualize unit responses via activation maximization. These synthetic feature visualizations are purported to provide humans with precise information about the image features that cause a unit to be activated - an advantage over other alternatives like strongly activating natural dataset samples. If humans indeed gain causal insight from visualizations, this should enable them to predict the effect of an intervention, such as how occluding a certain patch of the image (say, a dog's head) changes a unit's activation. Here, we test this hypothesis by asking humans to decide which of two square occlusions causes a larger change to a unit's activation. Both a large-scale crowdsourced experiment and measurements with experts show that on average the extremely activating feature visualizations by Olah et al. (2017) indeed help humans on this task (68±468 \pm 468±4% accuracy; baseline performance without any visualizations is 60±360 \pm 360±3%). However, they do not provide any substantial advantage over other visualizations (such as e.g. dataset samples), which yield similar performance (66±366\pm366±3% to 67±367 \pm367±3% accuracy). Taken together, we propose an objective psychophysical task to quantify the benefit of unit-level interpretability methods for humans, and find no evidence that a widely-used feature visualization method provides humans with better "causal understanding" of unit activations than simple alternative visualizations.

View on arXiv
Comments on this paper