Sparse but not Simpler: A Multi-Level Interpretability Analysis of Vision Transformers

• URL: http://arxiv.org/abs/2603.15919v2
• Date: Sun, 22 Mar 2026 19:33:42 GMT
• Title: Sparse but not Simpler: A Multi-Level Interpretability Analysis of Vision Transformers
• Abstract summary: We evaluate the relationship between weight sparsity and interpretability in Vision Transformers using DeiT-III B/16 models pruned with Wanda.<n>Our results reveal a clear structural effect but limited interpretability gains.<n>These findings suggest that structural sparsity alone does not reliably yield more interpretable vision models.
• Score: 5.6842776317118835
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Sparse neural networks are often hypothesized to be more interpretable than dense models, motivated by findings that weight sparsity can produce compact circuits in language models. However, it remains unclear whether structural sparsity itself leads to improved semantic interpretability. In this work, we systematically evaluate the relationship between weight sparsity and interpretability in Vision Transformers using DeiT-III B/16 models pruned with Wanda. To assess interpretability comprehensively, we introduce \textbf{IMPACT}, a multi-level framework that evaluates interpretability across four complementary levels: neurons, layer representations, task circuits, and model-level attribution. Layer representations are analyzed using BatchTopK sparse autoencoders, circuits are extracted via learnable node masking, and explanations are evaluated with transformer attribution using insertion and deletion metrics. Our results reveal a clear structural effect but limited interpretability gains. Sparse models produce circuits with approximately $2.5\times$ fewer edges than dense models, yet the fraction of active nodes remains similar or higher, indicating that pruning redistributes computation rather than isolating simpler functional modules. Consistent with this observation, sparse models show no systematic improvements in neuron-level selectivity, SAE feature interpretability, or attribution faithfulness. These findings suggest that structural sparsity alone does not reliably yield more interpretable vision models, highlighting the importance of evaluation frameworks that assess interpretability beyond circuit compactness.

Related papers

• Interpretability Without Tradeoffs: Disentangling Polysemanticity At Equal Predictive Performance [58.49623467782041]
  ELUDe is a method for improving the interpretability of deep neural networks.<n>It breaks latent representations into clear, inspectable sub-units that behave like interpretable features.<n>It requires no explicit training, no labels, and can be applied to pretrained models.
  arXiv  Detail & Related papers  (2026-05-29T13:39:09Z)
• Same Answer, Different Representations: Hidden instability in VLMs [65.36933543377346]
  We introduce a representation-aware and frequency-aware evaluation framework that measures internal embedding drift, spectral sensitivity, and structural smoothness.<n>We apply this framework to modern Vision Language Models (VLMs) across the SEEDBench, MMMU, and POPE datasets.
  arXiv  Detail & Related papers  (2026-02-06T12:24:26Z)
• I Predict Therefore I Am: Is Next Token Prediction Enough to Learn Human-Interpretable Concepts from Data? [76.15163242945813]
  Large language models (LLMs) have led many to conclude that they exhibit a form of intelligence.<n>We introduce a novel generative model that generates tokens on the basis of human-interpretable concepts represented as latent discrete variables.
  arXiv  Detail & Related papers  (2025-03-12T01:21:17Z)
• ULTra: Unveiling Latent Token Interpretability in Transformer-Based Understanding and Segmentation [14.84547724351634]
  We introduce ULTra, a framework for interpreting Transformer embeddings and uncovering meaningful semantic patterns within them.<n>We propose a self-supervised training approach that refines segmentation performance by learning an external transformation matrix without modifying the underlying model.<n>We validate ULTra for model interpretation on both synthetic and real-world scenarios, including Object Selection and interpretable text summarization.
  arXiv  Detail & Related papers  (2024-11-15T19:36:50Z)
• Improving Neuron-level Interpretability with White-box Language Models [11.898535906016907]
  We introduce a white-box transformer-like architecture named Coding RAte TransformEr (CRATE)<n>Our comprehensive experiments showcase significant improvements (up to 103% relative improvement) in neuron-level interpretability.<n>CRATE's increased interpretability comes from its enhanced ability to consistently and distinctively activate on relevant tokens.
  arXiv  Detail & Related papers  (2024-10-21T19:12:33Z)
• Improving Network Interpretability via Explanation Consistency Evaluation [56.14036428778861]
  We propose a framework that acquires more explainable activation heatmaps and simultaneously increase the model performance. Specifically, our framework introduces a new metric, i.e., explanation consistency, to reweight the training samples adaptively in model learning. Our framework then promotes the model learning by paying closer attention to those training samples with a high difference in explanations.
  arXiv  Detail & Related papers  (2024-08-08T17:20:08Z)
• Isomorphic Pruning for Vision Models [56.286064975443026]
  Structured pruning reduces the computational overhead of deep neural networks by removing redundant sub-structures. We present Isomorphic Pruning, a simple approach that demonstrates effectiveness across a range of network architectures.
  arXiv  Detail & Related papers  (2024-07-05T16:14:53Z)
• Explaining Text Similarity in Transformer Models [52.571158418102584]
  Recent advances in explainable AI have made it possible to mitigate limitations by leveraging improved explanations for Transformers. We use BiLRP, an extension developed for computing second-order explanations in bilinear similarity models, to investigate which feature interactions drive similarity in NLP models. Our findings contribute to a deeper understanding of different semantic similarity tasks and models, highlighting how novel explainable AI methods enable in-depth analyses and corpus-level insights.
  arXiv  Detail & Related papers  (2024-05-10T17:11:31Z)
• Interpretable time series neural representation for classification purposes [3.1201323892302444]
  The proposed model produces consistent, discrete, interpretable, and visualizable representations. The experiments show that the proposed model yields, on average better results than other interpretable approaches on multiple datasets.
  arXiv  Detail & Related papers  (2023-10-25T15:06:57Z)
• Improve Variational Autoencoder for Text Generationwith Discrete Latent Bottleneck [52.08901549360262]
  Variational autoencoders (VAEs) are essential tools in end-to-end representation learning. VAEs tend to ignore latent variables with a strong auto-regressive decoder. We propose a principled approach to enforce an implicit latent feature matching in a more compact latent space.
  arXiv  Detail & Related papers  (2020-04-22T14:41:37Z)

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.