AI RESEARCH
CauTion: Knowing When to Trust LLMs for Ensemble Causal Discovery
arXiv CS.LG
•
ArXi:2606.03602v1 Announce Type: new Causal discovery from observational data remains challenging due to the fundamental limitations of purely statistical methods, such as statistical distinguishability within equivalence classes and sensitivity to finite sample sizes. While large language models (LLMs) offer a promising source of domain knowledge to complement statistical inference, existing LLM-augmented methods are vulnerable to LLM errors and incur high token costs. Moreover, reliance on a single data-centric algorithm can make results sensitive to algorithm-specific biases.