When Bayesian neural networks remember too much
Two Bayesian continual-learning rules that differ by a single term, and what that term does to a posterior on a stream that will not sit still.
Writing
Notes on Bayesian learning, continual learning and the hardware it has to run on, plus the occasional portrait film.
2026
29 May Note
Active Continual Learning with Metaplastic Binary Bayesian Neural Networks
We are excited to share that our paper, Active Continual Learning with Metaplastic Binary Bayesian Neural Networks, has been accepted as a poster at ICML 2026 in Seoul, South Korea!
4 May Essay
A geometrical interpretation of variational methods in machine learning
Variational methods are often introduced through information theory: learning means finding an approximate distribution that remains close to a target Bayesian posterior.
2025
18 Dec Portrait
Itinérances - Ph.D Student Portrait - Lara Couronné
This portrait celebrates Lara Couronné, a brilliant scientist whose journey blends determination, passion, and creativity.
30 Oct Note
Bayesian Continual Learning in Nature Communications
I'm super happy to announce that Djohan Bonnet, Damien Querlioz and I, as well as co-authors, recently published a paper about Bayesian continual learning and forgetting in neural networks in Nature Communications.
2024
14 Nov Portrait
Itinérances - Ph.D Student Portrait - Kellian Cottart
Itinérances is a portrait of my journey as a Ph.D.
6 posts since 2024. New ones arrive in the RSS feed.