arXiv cs.AI· Chenxi Liu, Yongqiang Chen, Tongliang Liu, Mingming Gong, James Cheng, Bo Han, Kun Zhang·· 14 小时前AI 评分32
用大语言模型发现并推理隐藏世界中的因果关系：COAT 框架
Discovering and Reasoning of Causality in the Hidden World with Large Language Models
AI 导读
研究者提出 COAT（Causal representatiOn AssistanT）框架，利用 LLM 的世界知识为因果发现推荐可测量的高层变量，而非让 LLM 直接推断因果关系。
正文 · 原文
Abstract:Revealing hidden causal variables alongside the underlying causal mechanisms is essential to the development of science. Despite the progress in the past decades, existing practice in causal discovery (CD) heavily relies on high-quality measured variables, which are usually given by human experts. In fact, the lack of well-defined high-level variables behind unstructured data has been a longstanding roadblock to a broader real-world application of CD. This procedure can naturally benefit from an automated process that can suggest potential hidden variables in the system. Interestingly, Large language models (LLMs) are trained on massive observations of the world and have demonstrated great capability in processing unstructured data. To leverage the power of LLMs, we develop a new framework termed Causal representatiOn AssistanT (COAT) that incorporates the rich world knowledge of LLMs to propose useful measured variables for CD with respect to high-value target variables on their paired unstructured data. Instead of directly inferring causality with LLMs, COAT constructs feedback from intermediate CD results to LLMs to refine the proposed variables. Given the target variable and the paired unstructured data, we first develop COAT-MB that leverages the predictivity of the proposed variables to iteratively uncover the Markov Blanket of the target variable. Built upon COAT-MB, COAT-PAG further extends to uncover a more complete causal graph, i.e., Partial Ancestral Graph, by iterating over the target variables and actively seeking new high-level variables. Moreover, the reliable CD capabilities of COAT also extend the debiased causal inference to unstructured data by discovering an adjustment set. We establish theoretical guarantees for the CD results and verify their efficiency and reliability across realistic benchmarks and real-world case studies.
| Comments: | Accepted for publication in the Journal of Machine Learning Research. Extended journal version of our NeurIPS 2024 paper. Chenxi Liu and Yongqiang Chen contributed equally. 95 pages, 52 figures. Project page: this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Methodology (stat.ME) |
| Cite as: | arXiv:2402.03941 [cs.LG] |
| (or arXiv:2402.03941v4 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2402.03941 arXiv-issued DOI via DataCite |
Submission history
From: Chenxi Liu [view email]
[v1]
Tue, 6 Feb 2024 12:18:54 UTC (4,077 KB)
[v2]
Thu, 31 Oct 2024 12:27:30 UTC (3,518 KB)
[v3]
Mon, 13 Oct 2025 06:34:20 UTC (2,936 KB)
[v4]
Mon, 26 Oct 2026 04:29:18 UTC (2,773 KB)
来源：arXiv cs.AI · arxiv.org