Can GPT replace human raters of metaphors? Our new work with @vmangiaterra.bsky.social @cbarattieri.bsky.social @al-azary.bsky.social Across 687 Italian & English metaphors, GPT shows strong validity & reliability, but stumbles when meaning becomes sensorimotor ➡️ arxiv.org/abs/2512.12444 #NEPLab

Next Monday, the #IUSSLinguisticsSeminar will host Dr. Nadine Dijkstra, who will introduce us to the neural correlates of mental imagery, with a focus on the mechanisms allowing us to discern a mental image from a real visual percept. 🗓️June 16, 11-12.30am CEST Zoom link: tinyurl.com/keap95ks #NEPLab