TimeSAE: Sparse Decoding for Faithful Explanations of Black-Box Time Series Models

2026-01-14

AI Summary

This work addresses the limited generalization of existing black-box explanation methods for time series models under out-of-distribution scenarios. To overcome this challenge, the authors propose TimeSAE, a novel framework that introduces sparse autoencoders (SAEs) into time series interpretability for the first time. By integrating causal analysis, TimeSAE decouples and reconstructs the decision-making basis of target models to generate explanations that are both faithful and robust. Extensive experiments demonstrate that TimeSAE significantly outperforms state-of-the-art baselines across multiple synthetic and real-world datasets. Notably, it maintains high fidelity and reliability even under distribution shifts, thereby substantially enhancing the generalization capability of time series model explanations.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Computer Vision: Interpretability, Explainability, and Transparency

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

As black box models and pretrained models gain traction in time series applications, understanding and explaining their predictions becomes increasingly vital, especially in high-stakes domains where interpretability and trust are essential. However, most of the existing methods involve only in-distribution explanation, and do not generalize outside the training support, which requires the learning capability of generalization. In this work, we aim to provide a framework to explain black-box models for time series data through the dual lenses of Sparse Autoencoders (SAEs) and causality. We show that many current explanation methods are sensitive to distributional shifts, limiting their effectiveness in real-world scenarios. Building on the concept of Sparse Autoencoder, we introduce TimeSAE, a framework for black-box model explanation. We conduct extensive evaluations of TimeSAE on both synthetic and real-world time series datasets, comparing it to leading baselines. The results, supported by both quantitative metrics and qualitative insights, show that TimeSAE provides more faithful and robust explanations. Our code is available in an easy-to-use library TimeSAE-Lib: https://anonymous.4open.science/w/TimeSAE-571D/.

Problem

• time series
• black-box models
• model explanation
• distributional shift
• interpretability

Innovation

• Sparse Autoencoder
• Time Series Explanation
• Black-Box Interpretability
• Distributional Robustness
• Causal Explanation