Basile Confavreux

Excited to launch Principia, a nonprofit research organisation at the intersection of deep learning theory and AI safety.

Our goal is to develop theory for modern machine learning systems that can help us understand complex network behaviors, including those critical for AI safety and alignment.

1

Cosyne Viewing Parties

Visas, costs, care responsibilities, and environmental concerns all limit Cosyne attendance. Luckily, the talks are livestreamed; but watching alone is the high road to an aneurism. Hence: viewing parties! Gather regionally to watch Cosyne talks! More info: shorturl.at/3DHZX.

Thanks to amazing co-authors William Dorrell, Nishil Patel and supervisor @saxelab.bsky.social, and a special shoutout to @tpvogels.bsky.social!

We derive toy models that explain these seemingly accidental memories as input-dependent byproducts of movement in the manifold of stable weights - a form of memory - and a nice conceptual insight from meta-learned plasticity rules!

Find us at NeurIPS, Thur 4:30 pm #2115! We know networks have to be both plastic and stable but we're used to thinking about computations, such as memory, as additional requirements. Instead, we find that almost all stable & plastic networks display simple memory abilities.

Deep thanks to the brilliant team: @zoeharrington.bsky.social
@kaniamaciej.bsky.social
Poornima Ramesh, Anastasia Krouglova, @bozelosp.bsky.social @jakhmack.bsky.social @saxelab.bsky.social @ppjgoncalves.bsky.social @tpvogels.bsky.social

We’ve successfully automated part of the (neuro)scientific process. Now I may be out of a job. After a lot of iterations, here’s our framework for automatic discovery, built to embrace degeneracy and the realities of underconstrained modeling.

Memory by a thousand rules: Automated discovery of functional multi-type plasticity rules reveals variety & degeneracy at the heart of learning Synaptic plasticity is the basis of learning and memory, but the link between synaptic changes and neural function remains elusive. Here, we used automated search algorithms to obtain thousands of str...

We just pushed “Memory by a 1000 rules” onto bioRxiv, where we use clever #ML to find #plasticity quadruplets (EE, EI, IE, II) that learn basic stability in spiking nets. Why is it cool? We find 1000s!! of solutions, and they don’t just stabilise. They #memorise! www.biorxiv.org/content/10.1...