Unveiling Causal Reasoning in Large Language Models: Reality or Mirage?

2025-06-26

AI Summary

This study investigates whether large language models (LLMs) possess human-like level-2 causal reasoning—grounded in structural causal models and counterfactual reasoning—beyond superficial, correlation-based level-1 inference. Method: We propose G²-Reasoner, a framework integrating general-knowledge injection with goal-directed prompting to guide structured causal modeling. Leveraging the autoregressive nature of Transformer architectures, we introduce CausalProbe-2024, the first causal QA benchmark explicitly designed to evaluate novelty-aware and counterfactual reasoning. Contribution/Results: Experiments demonstrate that G²-Reasoner significantly outperforms baselines across cross-domain causal identification and counterfactual intervention inference. It validates a viable pathway from parameterized, experience-driven inference toward interpretable, generalizable level-2 causal reasoning. This work establishes a novel paradigm for advancing LLMs toward genuine causal intelligence.

Technology Category

Machine Learning: Causal LearningReasoning under Uncertainty: CausalityCognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingGraph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Causal reasoning capability is critical in advancing large language models (LLMs) toward strong artificial intelligence. While versatile LLMs appear to have demonstrated capabilities in understanding contextual causality and providing responses that obey the laws of causality, it remains unclear whether they perform genuine causal reasoning akin to humans. However, current evidence indicates the contrary. Specifically, LLMs are only capable of performing shallow (level-1) causal reasoning, primarily attributed to the causal knowledge embedded in their parameters, but they lack the capacity for genuine human-like (level-2) causal reasoning. To support this hypothesis, methodologically, we delve into the autoregression mechanism of transformer-based LLMs, revealing that it is not inherently causal. Empirically, we introduce a new causal Q&A benchmark called CausalProbe-2024, whose corpora are fresh and nearly unseen for the studied LLMs. The LLMs exhibit a significant performance drop on CausalProbe-2024 compared to earlier benchmarks, indicating the fact that they primarily engage in level-1 causal reasoning. To bridge the gap towards level-2 causal reasoning, we draw inspiration from the fact that human reasoning is usually facilitated by general knowledge and intended goals. We propose G^2-Reasoner, a method that incorporates general knowledge and goal-oriented prompts into LLMs' causal reasoning processes. Experiments demonstrate that G^2-Reasoner significantly enhances LLMs' causal reasoning capability, particularly in fresh and counterfactual contexts. This work sheds light on a new path for LLMs to advance towards genuine causal reasoning, going beyond level-1 and making strides towards level-2.

Problem

• Assessing if LLMs perform genuine human-like causal reasoning
• Identifying LLMs' limitation to shallow level-1 causal reasoning
• Proposing a method to enhance LLMs' causal reasoning capabilities

Innovation

• Transformer autoregression lacks inherent causality
• Introduces CausalProbe-2024 benchmark for testing
• Proposes G^2-Reasoner for enhanced causal reasoning