Local vs distributed representations: What is the right basis for interpretability?

• URL: http://arxiv.org/abs/2411.03993v1
• Date: Wed, 06 Nov 2024 15:34:57 GMT
• Title: Local vs distributed representations: What is the right basis for interpretability?
• Authors: Julien Colin, Lore Goetschalckx, Thomas Fel, Victor Boutin, Jay Gopal, Thomas Serre, Nuria Oliver,
• Abstract summary: We show that features obtained from sparse distributed representations are easier to interpret by human observers. Our results highlight that distributed representations constitute a superior basis for interpretability.
• Score: 19.50614357801837
• License:
• Abstract: Much of the research on the interpretability of deep neural networks has focused on studying the visual features that maximally activate individual neurons. However, recent work has cast doubts on the usefulness of such local representations for understanding the behavior of deep neural networks because individual neurons tend to respond to multiple unrelated visual patterns, a phenomenon referred to as "superposition". A promising alternative to disentangle these complex patterns is learning sparsely distributed vector representations from entire network layers, as the resulting basis vectors seemingly encode single identifiable visual patterns consistently. Thus, one would expect the resulting code to align better with human perceivable visual patterns, but supporting evidence remains, at best, anecdotal. To fill this gap, we conducted three large-scale psychophysics experiments collected from a pool of 560 participants. Our findings provide (i) strong evidence that features obtained from sparse distributed representations are easier to interpret by human observers and (ii) that this effect is more pronounced in the deepest layers of a neural network. Complementary analyses also reveal that (iii) features derived from sparse distributed representations contribute more to the model's decision. Overall, our results highlight that distributed representations constitute a superior basis for interpretability, underscoring a need for the field to move beyond the interpretation of local neural codes in favor of sparsely distributed ones.

Related papers

• Understanding polysemanticity in neural networks through coding theory [0.8702432681310401]
  We propose a novel practical approach to network interpretability and theoretical insights into polysemanticity and the density of codes. We show how random projections can reveal whether a network exhibits a smooth or non-differentiable code and hence how interpretable the code is. Our approach advances the pursuit of interpretability in neural networks, providing insights into their underlying structure and suggesting new avenues for circuit-level interpretability.
  arXiv  Detail & Related papers  (2024-01-31T16:31:54Z)
• Identifying Interpretable Visual Features in Artificial and Biological Neural Systems [3.604033202771937]
  Single neurons in neural networks are often interpretable in that they represent individual, intuitively meaningful features. Many neurons exhibit $textitmixed selectivity$, i.e., they represent multiple unrelated features. We propose an automated method for quantifying visual interpretability and an approach for finding meaningful directions in network activation space.
  arXiv  Detail & Related papers  (2023-10-17T17:41:28Z)
• Automated Natural Language Explanation of Deep Visual Neurons with Large Models [43.178568768100305]
  This paper proposes a novel post-hoc framework for generating semantic explanations of neurons with large foundation models. Our framework is designed to be compatible with various model architectures and datasets, automated and scalable neuron interpretation.
  arXiv  Detail & Related papers  (2023-10-16T17:04:51Z)
• DISCOVER: Making Vision Networks Interpretable via Competition and Dissection [11.028520416752325]
  This work contributes to post-hoc interpretability, and specifically Network Dissection. Our goal is to present a framework that makes it easier to discover the individual functionality of each neuron in a network trained on a vision task.
  arXiv  Detail & Related papers  (2023-10-07T21:57:23Z)
• Data-driven emergence of convolutional structure in neural networks [83.4920717252233]
  We show how fully-connected neural networks solving a discrimination task can learn a convolutional structure directly from their inputs. By carefully designing data models, we show that the emergence of this pattern is triggered by the non-Gaussian, higher-order local structure of the inputs.
  arXiv  Detail & Related papers  (2022-02-01T17:11:13Z)
• Redundant representations help generalization in wide neural networks [71.38860635025907]
  We study the last hidden layer representations of various state-of-the-art convolutional neural networks. We find that if the last hidden representation is wide enough, its neurons tend to split into groups that carry identical information, and differ from each other only by statistically independent noise.
  arXiv  Detail & Related papers  (2021-06-07T10:18:54Z)
• Adversarial Examples Detection with Bayesian Neural Network [57.185482121807716]
  We propose a new framework to detect adversarial examples motivated by the observations that random components can improve the smoothness of predictors. We propose a novel Bayesian adversarial example detector, short for BATer, to improve the performance of adversarial example detection.
  arXiv  Detail & Related papers  (2021-05-18T15:51:24Z)
• On 1/n neural representation and robustness [13.491651740693705]
  We show that imposing the experimentally observed structure on artificial neural networks makes them more robust to adversarial attacks. Our findings complement the existing theory relating wide neural networks to kernel methods.
  arXiv  Detail & Related papers  (2020-12-08T20:34:49Z)
• Domain Generalization for Medical Imaging Classification with Linear-Dependency Regularization [59.5104563755095]
  We introduce a simple but effective approach to improve the generalization capability of deep neural networks in the field of medical imaging classification. Motivated by the observation that the domain variability of the medical images is to some extent compact, we propose to learn a representative feature space through variational encoding.
  arXiv  Detail & Related papers  (2020-09-27T12:30:30Z)
• Towards Deeper Graph Neural Networks [63.46470695525957]
  Graph convolutions perform neighborhood aggregation and represent one of the most important graph operations. Several recent studies attribute this performance deterioration to the over-smoothing issue. We propose Deep Adaptive Graph Neural Network (DAGNN) to adaptively incorporate information from large receptive fields.
  arXiv  Detail & Related papers  (2020-07-18T01:11:14Z)
• Towards Understanding Hierarchical Learning: Benefits of Neural Representations [160.33479656108926]
  In this work, we demonstrate that intermediate neural representations add more flexibility to neural networks. We show that neural representation can achieve improved sample complexities compared with the raw input. Our results characterize when neural representations are beneficial, and may provide a new perspective on why depth is important in deep learning.
  arXiv  Detail & Related papers  (2020-06-24T02:44:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.