提出可一致的因果抽象网络，用数学结构提升模型可解释性。
Learning Consistent Causal Abstraction Networks
- 基于谱方法构建因果抽象网络，利用高斯结构和线性映射保证一致性。
- 在合成数据上成功恢复多种因果结构，性能优于现有方法。
- 适合研究因果建模、形式化推理与可解释AI的学者参考。
因果人工智能旨在通过结构因果模型（SCMs）提升AI的可解释性、可信度和鲁棒性。近期研究将网络层（sheaves）与余层（cosheaves）形式化为因果知识的表达方式。本文提出一致因果抽象网络（CAN），一种基于层论的框架：(i) SCMs为高斯分布；(ii) 限制映射为构造性线性因果抽象（CAs）的转置，符合语义嵌入原则；(iii) 边上的层（edge stalks）在排列等价下对应更细粒度SCMs的节点层（node stalks）。该问题被分解为边相关的局部黎曼优化问题，避免非凸目标。我们提出高效搜索算法，使用SPECTRAL——一种具有闭式更新的迭代方法，适用于正定与半正定协方差矩阵。在合成数据上的实验表明，该方法在因果抽象学习任务中表现优异，并能成功恢复多种复杂CAN结构。
原文摘要 · Abstract (English)
Causal artificial intelligence aims to enhance explainability, trustworthiness, and robustness in AI by leveraging structural causal models (SCMs). In this pursuit, recent advances formalize network sheaves and cosheaves of causal knowledge. Pushing in the same direction, we tackle the learning of consistent causal abstraction network (CAN), a sheaf-theoretic framework where (i) SCMs are Gaussian, (ii) restriction maps are transposes of constructive linear causal abstractions (CAs) adhering to the semantic embedding principle, and (iii) edge stalks correspond--up to permutation--to the node stalks of more detailed SCMs. Our problem formulation separates into edge-specific local Riemannian problems and avoids nonconvex objectives. We propose an efficient search procedure, solving the local problems with SPECTRAL, our iterative method with closed-form updates and suitable for positive definite and semidefinite covariance matrices. Experiments on synthetic data show competitive performance in the CA learning task, and successful recovery of diverse CAN structures.
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。