Formal Abductive Latent Explanations for Prototype-Based Networks

2025-11-20

AI Summary

Prototype networks possess inherent interpretability, yet their prototype-based explanations often suffer from ambiguity—distinct predictions may share identical explanations, severely undermining trustworthiness in safety-critical applications. To address this, we propose Abductive Latent Explanations (ALEs), the first framework integrating formal explainable AI with prototype networks. ALEs construct explanations in the latent space endowed with sufficient-condition guarantees, ensuring a theoretically sound one-to-one correspondence between explanations and predictions. Our method employs a scalable, SMT-solver-free algorithm that unifies counterfactual reasoning, causal inference, and constrained optimization to generate ALEs. Extensive evaluation on standard and fine-grained image classification benchmarks demonstrates substantial improvements in explanation consistency and reliability. This work provides the first formally guaranteed interpretability enhancement for prototype networks, advancing both theoretical rigor and practical applicability in trustworthy AI.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Philosophy and Ethics of AI: Accountability, Interpretability & Explainability

Application Category

• 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
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Case-based reasoning networks are machine-learning models that make predictions based on similarity between the input and prototypical parts of training samples, called prototypes. Such models are able to explain each decision by pointing to the prototypes that contributed the most to the final outcome. As the explanation is a core part of the prediction, they are often qualified as ``interpretable by design". While promising, we show that such explanations are sometimes misleading, which hampers their usefulness in safety-critical contexts. In particular, several instances may lead to different predictions and yet have the same explanation. Drawing inspiration from the field of formal eXplainable AI (FXAI), we propose Abductive Latent Explanations (ALEs), a formalism to express sufficient conditions on the intermediate (latent) representation of the instance that imply the prediction. Our approach combines the inherent interpretability of case-based reasoning models and the guarantees provided by formal XAI. We propose a solver-free and scalable algorithm for generating ALEs based on three distinct paradigms, compare them, and present the feasibility of our approach on diverse datasets for both standard and fine-grained image classification. The associated code can be found at https://github.com/julsoria/ale

Problem

• Addressing misleading explanations in prototype-based networks' predictions
• Ensuring sufficient conditions for predictions via latent representations
• Combining interpretable case-based reasoning with formal guarantees

Innovation

• Formal Abductive Latent Explanations for prototype networks
• Combines interpretable case-based reasoning with formal XAI guarantees
• Solver-free algorithm generates sufficient latent conditions for predictions