SOURCE-LINKED INTELLIGENCE

When and Why LLM Causal Priors Help: Closed-Loop Prior Selection for Amortized Causal Inference

arXiv · AI, language, vision and robotics · article · Sep 7, 2026 · UTC

Causal effect estimation asks how an outcome would change under an intervention, and medicine, economics, and public policy all treat it as a foundational task. Prior-data fitted networks (PFNs) amortize the task: a model trained on large numbers of programmatically generated synthetic causal tasks reads a new problem's observational data into context and returns an interventional-effect estimate in a single forward pass. The capability of such models is largely determined by the synthetic training prior, which is currently designed by hand, a bottleneck acknowledged by both Do-PFN and CausalP

Read original source ↗ Open in workspace

recordType

paper

region

Global