MONOLO Logo
Workshop

MANOLO Workshop on Growing Neural Networks

07/12/26 - 07/12/26
Université Paris 1 Panthéon-Sorbonne, Centre La Chapelle, Paris
This workshop brings together researchers across communities to explore how optimisation should adapt when neural network architectures change during training, through talks, a poster session, and a panel discussion.

The workshop focuses on methods that change the network structure during training: growing neurons and layers, progressive scaling of transformers, dynamic sparse training, and applications to continual and reinforcement learning. They all share one open question: once the architecture changes, how should optimisation continue? This work is spread across several communities with little contact between them, and the aim of the day is to bring them into one room.

The programme includes talks from Kaitlin Maile (Google), Manon Verbockhaven (ENS Lyon), Pierre Wolinski (Université Paris Dauphine–PSL), and Rupert Mitchell (TU Darmstadt), with further speakers to be announced.

There will also be a poster session and a panel discussion.

ELLIS Edge Newsletter
Join the 6,000+ people who get the monthly newsletter filled with the latest news, jobs, events and insights from the ELLIS Network.