Failure Modes of LLMs for Causal Reasoning on Narratives

Khurram Yamin, Shantanu Gupta, Gaurav Rohit Ghosal, Zachary Chase Lipton, Bryan Wilder

Keywords: Causal Inference, Large Language Models, Reasoning, Narratives

TL;DR: In this paper, we examine the failure Modes of LLMs for causal reasoning on narratives and the unreliable shortcuts LLMs take to make causal inferences..

Abstract:

In this work, we investigate the causal reasoning abilities of large language models (LLMs) through the representative problem of inferring causal relationships from narratives. We find that even state of the art language models rely heavily on unreliable shortcuts, both in terms of the narrative presentation and their parametric knowledge. For example, LLMs tend to determine causal relationships based on the temporal ordering of events (i.e., earlier events cause later ones), resulting in lower performance whenever events are not narrated in their exact causal order. Similarly, we demonstrate that LLMs struggle with long-term causal reasoning — they often fail when the narratives are longer and contain many events. As an additional failure mode, we show LLMs appear to heavily rely on their parametric knowledge at the expense of reasoning over the provided narrative. This degrades their abilities whenever the narrative opposes parametric knowledge. We extensively validate these failure modes through carefully controlled synthetic experiments, as well as evaluations on real-world narratives. Finally, we observe that explicitly generating a causal graph generally improves performance
while naive chain-of-thought is ineffective. Collectively, our results distill precise failure modes of current state-of-the art models and can pave the way for future techniques to enhance causal reasoning in LLMs.

Primary Area: foundation or frontier models, including LLMs

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 12689