让神经网络提前输出更可信：用注意力一致性提升解释性
Attention Consistency Regularization for Interpretable Early-Exit Neural Networks
- 通过注意力对齐损失，使早退出层与最终输出保持一致
- 在保持98.97%准确率的同时提速1.97倍，注意力一致性提升18.5%
- 适合需要可解释性的低资源部署场景
早退出神经网络通过在中间层提前做出预测，实现自适应推理以降低计算开销。然而，早退出常缺乏可解释性，且关注特征与深层网络不同，影响可信度。本文提出解释引导训练（EGT），一种多目标框架，通过基于注意力的正则化提升早退出网络的可解释性与一致性。EGT引入注意力一致性损失，使早退出注意力图与最终输出对齐，并通过加权损失联合优化分类准确率与注意力一致性。在真实图像分类数据集上的实验表明，EGT在保持98.97%整体准确率（与基线相当）的同时，实现1.97倍的推理加速，注意力一致性相比基线提升最高达18.5%。该方法在所有退出点均提供更可解释、一致的解释，使早退出网络更适合资源受限环境中的可解释人工智能应用。
原文摘要 · Abstract (English)
Early-exit neural networks enable adaptive inference by allowing predictions at intermediate layers, reducing computational cost. However, early exits often lack interpretability and may focus on different features than deeper layers, limiting trust and explainability. This paper presents Explanation-Guided Training (EGT), a multi-objective framework that improves interpretability and consistency in early-exit networks through attention-based regularization. EGT introduces an attention consistency loss that aligns early-exit attention maps with the final exit. The framework jointly optimizes classification accuracy and attention consistency through a weighted combination of losses. Experiments on a real-world image classification dataset demonstrate that EGT achieves up to 98.97% overall accuracy (matching baseline performance) with a 1.97x inference speedup through early exits, while improving attention consistency by up to 18.5% compared to baseline models. The proposed method provides more interpretable and consistent explanations across all exit points, making early-exit networks more suitable for explainable AI applications in resource-constrained environments.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。