AI‑crafted stories expose what language patches of brain actually detect
Microsoft Research’s generative causal testing (GCT) turns opaque language‑model predictions into short, testable hypotheses about cortical patches—like “food preparation” or “clock times”. By feeding LLM‑written stories to fMRI subjects, researchers prove the targeted region lights up only when the explanation is correct, closing the gap between prediction and neuroscientific understanding.