Fast Explanations via Policy Gradient-Optimized Explainer

2024-05-29

AI Summary

To address the practical deployment challenges of model interpretation methods—namely, high query overhead and strong model dependency—this paper proposes a model-agnostic, end-to-end differentiable attribution framework. Methodologically, it reformulates explanation generation as a policy learning problem, representing attributions via a learnable probability distribution and optimizing inference efficiency through policy gradient methods. Crucially, the approach requires neither access to model internals nor gradients, nor does it rely on extensive forward queries. Evaluated on image and text classification tasks, it achieves over 97% faster inference and reduces memory consumption by 70%, while preserving high fidelity and cross-model generalizability. These results demonstrate a significant breakthrough in resolving the long-standing trade-off between computational efficiency and methodological generality in attribution-based explanation.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Transparent, Interpretable, Explainable ML
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The challenge of delivering efficient explanations is a critical barrier that prevents the adoption of model explanations in real-world applications. Existing approaches often depend on extensive model queries for sample-level explanations or rely on expert's knowledge of specific model structures that trade general applicability for efficiency. To address these limitations, this paper introduces a novel framework Fast Explanation (FEX) that represents attribution-based explanations via probability distributions, which are optimized by leveraging the policy gradient method. The proposed framework offers a robust, scalable solution for real-time, large-scale model explanations, bridging the gap between efficiency and applicability. We validate our framework on image and text classification tasks and the experiments demonstrate that our method reduces inference time by over 97% and memory usage by 70% compared to traditional model-agnostic approaches while maintaining high-quality explanations and broad applicability.

Problem

• Model Interpretability
• Efficiency Improvement
• Query Reduction

Innovation

• FEX
• Efficient Interpretability
• Policy Gradient Optimization