Causal Discovery with Language Models as Imperfect Experts
Stephanie Long Alexandre Piché Valentina Zantedeschi Tibor Schuster Alexandre Drouin
Abstract
Understanding the causal relationships that underlie a system is a fundamental prerequisite to accurate decision-making. In this work, we explore how expert knowledge can be used to improve the data-driven identification of causal graphs, beyond Markov equivalence classes. In doing so, we consider a setting where we can query an expert about the orientation of causal relationships between variables, but where the expert may provide erroneous information. We propose strategies for amending such expert knowledge based on consistency properties, e.g., acyclicity and conditional independencies in the equivalence class. We then report a case study, on real data, where a large language model is used as an imperfect expert.
中文速览
从观测数据中还原因果图时，算法通常只能确定一个"等价类"而无法区分其中哪个图才是真正的因果结构，引入领域专家的判断可以进一步缩小这个等价类，但现实中专家（包括大语言模型）的判断并不总是正确。本文将这一问题建模为一个优化目标：在允许专家犯错的前提下，尽可能缩小马尔可夫等价类（Markov equivalence class），同时以高概率保证真实因果图仍被包含其中。为此，作者提出了一种基于贝叶斯推断的贪心策略，利用无环性和条件独立性等一致性约束来识别并过滤专家的错误判断，从而在可控风险下逐步锁定边的方向。在多个真实贝叶斯网络上的实验表明，该方法在专家错误率适中时能显著压缩等价类大小并降低结构误差，但以GPT-3.5作为专家时效果参差不齐，揭示了大语言模型在细粒度因果推断任务中的局限性。这项工作为在不完美的外部知识源（如大语言模型）与数据驱动因果发现之间架起桥梁提供了一套系统性框架，对推动因果发现的实用化落地具有重要意义。
原文 arXiv:2307.02390;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2307.02390v1