Tag
This paper introduces Causal Foundation Models, which use pretrained neural networks to estimate causal effects on new datasets via in-context learning without fine-tuning, providing a practical guide to this emerging field.