Matteo Priorelli

I am a PhD student at Sapienza University of Rome, working on neurocomputational modeling and supervised by Giovanni Pezzulo, in collaboration with the Institute of Cognitive Sciences and Technologies (National Reseach Council of Italy).
My interests lie at the intersection of theoretical neuroscience and machine learning. My goal is to understand how cognition emerges from neural structures and how it can be replicated to build intelligent agents.
My current research is based on predictive coding, active inference, and Bayesian approaches. My latest projects contain some proposals on dynamic goal-directed behavior in active inference.
These are Daniel from The Karate Kid and Neo from The Matrix:


latest posts
selected publications
- Deep kinematic inference affords efficient and scalable control of bodily movementsPNAS, 2023
- Dynamic Inference by Model ReductionbioRxiv, 2023
- Deep hybrid models: infer and plan in a dynamic worldEntropy, 2025
- Embodied decisions as active inferencebioRxiv, Jun 2024