Invariant Rationalization

摆脱 MMI：通过探测输入利用情况开辟理性化新前沿

Breaking Free from MMI: A New Frontier in Rationalization by Probing Input Utilization

本文指出最大互信息准则存在边际收益递减问题，提出利用输入表示范数作为替代目标，通过匹配权重矩阵能力空间识别网络真正能利用的理性片段，并在多个数据集上验证其优于 MMI 及部分 LLM。

MMI 准则对于可解释性是否必要？将非因果特征退化为纯噪声以实现自理性化

Is the MMI Criterion Necessary for Interpretability? Degenerating Non-causal Features to Plain Noise for Self-Rationalization

本文提出最大化剩余差异（MRD）准则，将非因果虚假特征视为纯噪声，以替代传统 MMI 准则进行理性提取，在六个数据集上相比近期 MMI 变体将理性质量提升最高 10.4%。

细粒度图理性化

Fine-grained Graph Rationalization

本文提出细粒度图理性化（FIG），利用自注意力机制实现节点级和虚拟节点级干预，在7个数据集上超越13种基线方法，显著提升图理性化性能。

面向图泛化的协同分类与合理化方法

Cooperative Classification and Rationalization for Graph Generalization

针对图神经网络分布外泛化难的问题，提出由分类和合理化模块组成的C2R方法，通过环境条件生成、依据子图分离、知识蒸馏对齐及多环境协同学习提升泛化能力，实验验证了其有效性。

Max-Margin Invariant Features from Transformed Unlabeled Data

The study of representations invariant to common transformations of the data is important to learning. Most techniques have focused on local approximate invariance implemented within expensive optimization frameworks lacking explicit theoretical guarantees. In this paper, we study kernels that are invariant to a unitary group while having theoretical guarantees in addressing the important practical issue of unavailability of transformed versions of labelled data. A problem we call the Unlabeled Transformation Problem which is a special form of semi-supervised learning and one-shot learning. We…

Discovering Invariant Rationales for Graph Neural Networks

Intrinsic interpretability of graph neural networks (GNNs) is to find a small subset of the input graph's features -- rationale -- which guides the model prediction. Unfortunately, the leading rationalization models often rely on data biases, especially shortcut features, to compose rationales and make predictions without probing the critical and causal patterns. Moreover, such data biases easily change outside the training distribution. As a result, these models suffer from a huge drop in interpretability and predictive performance on out-of-distribution data. In this work, we propose a new s…