CARE: Turning LLMs Into Causal Reasoning Expert

TL;DR

This paper introduces CARE, a framework that fine-tunes LLMs to improve their causal reasoning by effectively integrating outputs from established causal discovery algorithms, surpassing existing methods in accuracy.

Contribution

The paper presents a novel fine-tuning approach that enhances LLMs' causal reasoning by teaching them to utilize external causal discovery outputs, addressing their inherent limitations.

Findings

Fine-tuned Qwen2.5-1.5B outperforms larger models and traditional algorithms.

Prompting with causal discovery outputs initially decreases LLM performance.

CARE significantly improves LLMs' ability to perform causal reasoning.

Abstract

Large language models (LLMs) have recently demonstrated impressive capabilities across a range of reasoning and generation tasks. However, research studies have shown that LLMs lack the ability to identify causal relationships, a fundamental cornerstone of human intelligence. We first conduct an exploratory investigation of LLMs' behavior when asked to perform a causal-discovery task and find that they mostly rely on the semantic meaning of variable names, ignoring the observation data. This is unsurprising, given that LLMs were never trained to process structural datasets. To first tackle this challenge, we prompt the LLMs with the outputs of established causal discovery algorithms designed for observational datasets. These algorithm outputs effectively serve as the sufficient statistics of the observation data. However, quite surprisingly, we find that prompting the LLMs with these sufficient statistics decreases the LLMs' performance in causal discovery. To address this current limitation, we propose CARE, a framework that enhances LLMs' causal-reasoning ability by teaching them to effectively utilize the outputs of established causal-discovery algorithms through supervised fine-tuning. Experimental results show that a finetuned Qwen2.5-1.5B model produced by CARE significantly outperforms both traditional causal-discovery algorithms and state-of-the-art LLMs with over a thousand times more parameters, demonstrating effective utilization of its own knowledge and the external algorithmic clues.