Seeing Through Circuits: Faithful Mechanistic Interpretability for Vision Transformers

2026-04-15

AI Summary

Existing interpretability methods for vision models predominantly focus on neuron activations, lacking a mechanistic understanding of how information propagates through the model. This work proposes Visual Circuit Discovery (Vi-CD), the first method to introduce edge-level mechanistic circuit analysis into Vision Transformers. By constructing an edge-based computational graph, Vi-CD automatically identifies class-specific information pathways relevant to particular tasks. The approach not only uncovers the internal information routing mechanisms of vision models but also successfully locates adversarial circuits in CLIP that underlie typographic attacks. Furthermore, targeted interventions on these discovered circuits effectively mitigate harmful behaviors, enabling transparent and controllable manipulation of visual model decision processes.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Probabilistic Circuits and Graphical Models
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Transparency of neural networks' internal reasoning is at the heart of interpretability research, adding to trust, safety, and understanding of these models. The field of mechanistic interpretability has recently focused on studying task-specific computational graphs, defined by connections (edges) between model components. Such edge-based circuits have been defined in the context of large language models, yet vision-based approaches so far only consider neuron-based circuits. These tell which information is encoded, but not how it is routed through the complex wiring of a neural network. In this work, we investigate whether useful mechanistic circuits can be identified through computational graphs in vision transformers. We propose an effective method for Automatic Visual Circuit Discovery (Vi-CD) that recovers class-specific circuits for classification, identifies circuits underlying typographic attacks in CLIP, and discovers circuits that lend themselves for steering to correct harmful model behavior. Overall, we find that insightful and actionable edge-based circuits can be recovered from vision transformers, adding transparency to the internal computations of these models.

Problem

• mechanistic interpretability
• vision transformers
• computational graphs
• edge-based circuits
• model transparency

Innovation

• mechanistic interpretability
• vision transformers
• computational graphs
• circuit discovery
• edge-based circuits