The Magic of IF: Investigating Causal Reasoning Abilities in Large Language Models of Code
This addresses the challenge of improving causal reasoning in LLMs for AI applications, though it is incremental as it builds on existing models with code-specific prompts.
The study investigated whether Code-LLMs have better causal reasoning abilities than text-only LLMs, finding that Code-LLMs with code prompts significantly outperform them in tasks like abductive and counterfactual reasoning.
Causal reasoning, the ability to identify cause-and-effect relationship, is crucial in human thinking. Although large language models (LLMs) succeed in many NLP tasks, it is still challenging for them to conduct complex causal reasoning like abductive reasoning and counterfactual reasoning. Given the fact that programming code may express causal relations more often and explicitly with conditional statements like ``if``, we want to explore whether Code-LLMs acquire better causal reasoning abilities. Our experiments show that compared to text-only LLMs, Code-LLMs with code prompts are significantly better in causal reasoning. We further intervene on the prompts from different aspects, and discover that the programming structure is crucial in code prompt design, while Code-LLMs are robust towards format perturbations.