让概念模型真正懂因果，提升可解释性和泛化能力。
Causally Reliable Concept Bottleneck Models
- 用真实因果结构约束概念瓶颈，实现可信推理。
- 在多个任务上比传统模型更准确响应干预措施。
- 适合需要可解释性与公平性的实际应用者。
基于概念的模型是深度学习中一种新兴范式，通过人类可理解变量约束推理过程，增强可解释性与人机交互。然而，这类架构与主流黑箱神经网络一样，未能反映数据背后真实因果机制，导致难以支持因果推理、限制分布外泛化，并阻碍公平性约束的实施。为此，我们提出因果可靠的概念瓶颈模型（C$^2$BMs），其概念瓶颈结构基于对现实世界因果机制的建模。我们还设计了一套管道，可从观测数据和非结构化背景知识（如科学文献）中自动学习该结构。实验表明，相较于标准黑箱模型与概念模型，C$^2$BMs在保持准确性的同时，具备更强可解释性、因果可靠性，并能更好响应干预操作。
原文摘要 · Abstract (English)
Concept-based models are an emerging paradigm in deep learning that constrains the inference process to operate through human-interpretable variables, facilitating explainability and human interaction. However, these architectures, on par with popular opaque neural models, fail to account for the true causal mechanisms underlying the target phenomena represented in the data. This hampers their ability to support causal reasoning tasks, limits out-of-distribution generalization, and hinders the implementation of fairness constraints. To overcome these issues, we propose Causally reliable Concept Bottleneck Models (C$^2$BMs), a class of concept-based architectures that enforce reasoning through a bottleneck of concepts structured according to a model of the real-world causal mechanisms. We also introduce a pipeline to automatically learn this structure from observational data and unstructured background knowledge (e.g., scientific literature). Experimental evidence suggests that C$^2$BMs are more interpretable, causally reliable, and improve responsiveness to interventions w.r.t. standard opaque and concept-based models, while maintaining their accuracy.
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。