Faster Verified Explanations for Neural Networks

2025-11-28

AI Summary

Formal verification-based explanations for neural networks suffer from prohibitive computational overhead and poor scalability on large-scale models. Method: This paper proposes FaVeX, a novel algorithm that (i) introduces a verifier-optimal robust explanation framework explicitly modeling verifier incompleteness, and (ii) integrates dynamic batching with query information reuse to jointly accelerate both invariance proof and critical feature identification. By unifying formal verification with efficient inference, FaVeX enables scalable, semantically meaningful formal explanation generation—even on deep networks with hundreds of thousands of nonlinear activation units. Contribution/Results: Experiments demonstrate that FaVeX achieves order-of-magnitude speedups over state-of-the-art baselines while significantly strengthening robustness guarantees. It establishes a new paradigm for high-assurance, interpretable AI systems.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Humans and AI: Explainable AI (XAI) for Human Understanding

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Verified explanations are a theoretically-principled way to explain the decisions taken by neural networks, which are otherwise black-box in nature. However, these techniques face significant scalability challenges, as they require multiple calls to neural network verifiers, each of them with an exponential worst-case complexity. We present FaVeX, a novel algorithm to compute verified explanations. FaVeX accelerates the computation by dynamically combining batch and sequential processing of input features, and by reusing information from previous queries, both when proving invariances with respect to certain input features, and when searching for feature assignments altering the prediction. Furthermore, we present a novel and hierarchical definition of verified explanations, termed verifier-optimal robust explanations, that explicitly factors the incompleteness of network verifiers within the explanation. Our comprehensive experimental evaluation demonstrates the superior scalability of both FaVeX, and of verifier-optimal robust explanations, which together can produce meaningful formal explanation on networks with hundreds of thousands of non-linear activations.

Problem

• Accelerates verified explanations for neural networks
• Addresses scalability of neural network verifiers
• Introduces hierarchical verifier-optimal robust explanations

Innovation

• Dynamic batch and sequential processing of features
• Reusing information from previous verification queries
• Hierarchical verifier-optimal robust explanations definition