CausalArena: Benchmarking Causal Discovery in the Foundation Model Era

Researchers introduced CausalArena, a unified and evolvable benchmark for causal discovery, detailed in arXiv paper 2609.11897v1. CausalArena combines synthetic structural causal models (SCMs), semantic operational SCMs, formula-grounded SCMs, and public real-world datasets under a common protocol. Experiments across classical, neural, and pretrained methods showed substantial ranking shifts across SCM families and protocols, indicating that strong performance in one benchmark regime does not reliably transfer to others and that pretraining-evaluation overlap complicates assessment of causal discovery foundation models (CDFMs).

arXiv:2609.11897v1 Announce Type: new Abstract: Causal discovery aims to uncover causal structures from data and is fundamental to scientific reasoning and intervention-based decision making. Its evaluation relies heavily on structural causal models SCMs , which specify a causal graph together with the mechanisms that generate data, yet existing studies differ substantially in graph families, mechanisms, and evaluation protocols. The emergence of causal discovery foundation models CDFMs further complicates evaluation: performance may reflect not only causal discovery ability, but also overlap between pretraining environments and test SCMs, making results on fixed synthetic benchmarks difficult to interpret. We introduce CausalArena, a unified and evolvable benchmark for causal discovery under a common protocol. Synthetic SCMs supply controlled breadth over structures and mechanisms; semantic operational SCMs provide human-auditable, semantically grounded environments beyond standard synthetic generators; and formula-grounded SCMs test discovery under explicit scientific mechanisms. Public real-world datasets provide an additional external-validity check. Experiments across classical, neural, and pretrained methods reveal substantial ranking shifts across SCM families and protocols, showing that strong performance in one benchmark regime does not reliably transfer to others. These results highlight benchmark diversity and pretraining--evaluation overlap as central challenges for evaluating causal discovery in the foundation model era.