Causal Discovery Bench: Can LLMs Autonomously Conduct Effective Experiments?

Causal-Discovery-Bench is a verifiable sandbox to evaluate sequential experimental design and causal discovery without domain-semantic cues. Across 3,802 episodes with 13 models from 5 vendors, agent performance shows a four-level difficulty hierarchy. This hierarchy spans single-experiment design, sequential exploration, resource-aware experimentation, and autonomous resolution of conflicting evidence.