VITAL: More Understandable Feature Visualization through Distribution Alignment and Relevant Information Flow

• URL: http://arxiv.org/abs/2503.22399v1
• Date: Fri, 28 Mar 2025 13:08:18 GMT
• Title: VITAL: More Understandable Feature Visualization through Distribution Alignment and Relevant Information Flow
• Authors: Ada Gorgun, Bernt Schiele, Jonas Fischer,
• Abstract summary: Feature visualization (FV) is a powerful tool to decode what information neurons are responding to.<n>We propose to guide FV through statistics of prototypical image features combined with measures of relevant network flow to generate images.<n>Our approach yields human-understandable visualizations that both qualitatively and quantitatively improve over state-of-the-art FVs.
• Score: 57.96482272333649
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Neural networks are widely adopted to solve complex and challenging tasks. Especially in high-stakes decision-making, understanding their reasoning process is crucial, yet proves challenging for modern deep networks. Feature visualization (FV) is a powerful tool to decode what information neurons are responding to and hence to better understand the reasoning behind such networks. In particular, in FV we generate human-understandable images that reflect the information detected by neurons of interest. However, current methods often yield unrecognizable visualizations, exhibiting repetitive patterns and visual artifacts that are hard to understand for a human. To address these problems, we propose to guide FV through statistics of real image features combined with measures of relevant network flow to generate prototypical images. Our approach yields human-understandable visualizations that both qualitatively and quantitatively improve over state-of-the-art FVs across various architectures. As such, it can be used to decode which information the network uses, complementing mechanistic circuits that identify where it is encoded. Code is available at: https://github.com/adagorgun/VITAL

Related papers

• Generating visual explanations from deep networks using implicit neural representations [0.6056822594090163]
  In this work, we demonstrate that implicit neural representations (INRs) constitute a good framework for generating visual explanations. We present an iterative INR-based method that can be used to generate multiple non-overlapping attribution masks for the same image.
  arXiv  Detail & Related papers  (2025-01-20T23:17:57Z)
• Saliency Suppressed, Semantics Surfaced: Visual Transformations in Neural Networks and the Brain [0.0]
  We take inspiration from neuroscience to shed light on how neural networks encode information at low (visual saliency) and high (semantic similarity) levels of abstraction. We find that ResNets are more sensitive to saliency information than ViTs, when trained with object classification objectives. We show that semantic encoding is a key factor in aligning AI with human visual perception, while saliency suppression is a non-brain-like strategy.
  arXiv  Detail & Related papers  (2024-04-29T15:05:42Z)
• Unleashing the Power of Depth and Pose Estimation Neural Networks by Designing Compatible Endoscopic Images [12.412060445862842]
  We conduct a detail analysis of the properties of endoscopic images and improve the compatibility of images and neural networks. First, we introcude the Mask Image Modelling (MIM) module, which inputs partial image information instead of complete image information. Second, we propose a lightweight neural network to enhance the endoscopic images, to explicitly improve the compatibility between images and neural networks.
  arXiv  Detail & Related papers  (2023-09-14T02:19:38Z)
• Don't trust your eyes: on the (un)reliability of feature visualizations [25.018840023636546]
  We show how to trick feature visualizations into showing arbitrary patterns that are completely disconnected from normal network behavior on natural input. We then provide evidence for a similar phenomenon occurring in standard, unmanipulated networks. This can be used as a sanity check for feature visualizations.
  arXiv  Detail & Related papers  (2023-06-07T18:31:39Z)
• Synergistic information supports modality integration and flexible learning in neural networks solving multiple tasks [107.8565143456161]
  We investigate the information processing strategies adopted by simple artificial neural networks performing a variety of cognitive tasks. Results show that synergy increases as neural networks learn multiple diverse tasks. randomly turning off neurons during training through dropout increases network redundancy, corresponding to an increase in robustness.
  arXiv  Detail & Related papers  (2022-10-06T15:36:27Z)
• Learning Enriched Features for Fast Image Restoration and Enhancement [166.17296369600774]
  This paper presents a holistic goal of maintaining spatially-precise high-resolution representations through the entire network. We learn an enriched set of features that combines contextual information from multiple scales, while simultaneously preserving the high-resolution spatial details. Our approach achieves state-of-the-art results for a variety of image processing tasks, including defocus deblurring, image denoising, super-resolution, and image enhancement.
  arXiv  Detail & Related papers  (2022-04-19T17:59:45Z)
• A neural anisotropic view of underspecification in deep learning [60.119023683371736]
  We show that the way neural networks handle the underspecification of problems is highly dependent on the data representation. Our results highlight that understanding the architectural inductive bias in deep learning is fundamental to address the fairness, robustness, and generalization of these systems.
  arXiv  Detail & Related papers  (2021-04-29T14:31:09Z)
• Proactive Pseudo-Intervention: Causally Informed Contrastive Learning For Interpretable Vision Models [103.64435911083432]
  We present a novel contrastive learning strategy called it Proactive Pseudo-Intervention (PPI) PPI leverages proactive interventions to guard against image features with no causal relevance. We also devise a novel causally informed salience mapping module to identify key image pixels to intervene, and show it greatly facilitates model interpretability.
  arXiv  Detail & Related papers  (2020-12-06T20:30:26Z)
• Ventral-Dorsal Neural Networks: Object Detection via Selective Attention [51.79577908317031]
  We propose a new framework called Ventral-Dorsal Networks (VDNets) Inspired by the structure of the human visual system, we propose the integration of a "Ventral Network" and a "Dorsal Network" Our experimental results reveal that the proposed method outperforms state-of-the-art object detection approaches.
  arXiv  Detail & Related papers  (2020-05-15T23:57:36Z)

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.