About me
Hi! Iβm Francesco.
Iβm currently a PhD student in Machine Learning and Theoretical Neuroscience at the University of Sussex, supervised by Christopher Buckley and Anil Seth. My research has so far focused on brain-inspired learning algorithms for training neural networks as an alternative to standard backpropagation. More broadly, I am interested in energy-based models, the optimisation dynamics of neural networks, and all things AI-related.
Previously, I interned as an Applied Scientist at Amazon, helping to improve their model forecasts to deliver packages throughout Europe more efficiently. During my undergrad, I was also a research assistant in the lab of Devin Terhune, using machine learning tools to differentiate the subjective experiences associated with different psychedelic drugs.
Outside of work, I enjoy swimming, running, and playing football, and am always up for a physical challenge!
News
π¨ Paper republished in Journal of Statistical Mechanics
July 2025
A slightly updated version of our NeuIPS 2024 paper on the geometry of the energy landscape of predictive coding networks has just been republished here in the Journal of Statistical Mechanics: Theory and Experiment as part of a Special Issue on Machine Learning 2025.
π₯ Training 100+ layer networks with a local algorithm
May 2025
We introduce ΞΌPC, a reparameterisation of predictive coding that allows stable training of 100+ layer residual networks with zero-shot hyperparameter transfer. See this post for a summary. Code here.
π»β Introducing JPC
December 2024
βοΈ Our lab just released JPC , a JAX library for training neural networks with Predictive Coding. See this pre-print for more details.
π¨π¦ Paper accepted at NeurIPS 2024
September 2024
Our paper Only Strict Saddles in the Energy Landscape of Predictive Coding Networks? has been accepted at NeurIPS 2024! See this blog post for the key takeaways.
πͺπΈ Applied Scientist Intern at Amazon
April 2024
I recently completed an internship as Applied Scientist at Amazon, in Barcelona. I wrote about my experience here if youβre interested.
π Best Workshop Paper Award at ICML 2023
June 2023
I will be at ICML 2023 in Hawaii to present my work on Understanding Predictive Coding as a Second-Order Trust-Region Method, which won a Best Paper Award at the Workshop on Localized Learning. You can rewatch my talk here or read my blog post for the key ideas.
Honoured to share that my first PhD paper "Understanding Predictive Coding as a Second-Order Trust-Region Method" (https://t.co/xP3m0aZS99) won a Best Paper Award at the ICML 2023 Workshop on Localized Learning! π #icml2023
— Francesco Innocenti (@InnocFrancesco) July 17, 2023
If this piques your interest, read on 1/14ππ§΅