How can scientists determine cause and effect when experiments are impossible? A new framework published in The Innovation uses delay embedding entropy to infer intervention-level causality directly from observational data, eliminating hidden confounders in neural circuits and complex systems.