Nicolas Zucchet

PhD student at ETH Zürich

prof_pic.jpg

I am a third-year PhD student at ETH Zürich, under the supervision of João Sacramento and Angelika Steger. I earned my Master’s degree in Computer Science from ETH Zürich in 2021, following a Bachelor’s degree in Applied Mathematics from École Polytechnique in Palaiseau, France, in 2019.

My research interests lie at the intersection of Artificial Intelligence and Neuroscience, focusing on understanding how brains learn. I aim to create algorithms inspired by deep learning that perform well while fitting within the brain’s physical limits. For example, I’m interested in developing online learning rules for neural networks that do not require any future information. My theoretical background fuels my curiosity about the fundamental principles underlying learning. This drives my work in deep learning, where I study the optimization properties of state-of-the-art architectures, as well as the kind of computations they can perform.

Outside research, I enjoy long-distance cycling to appreciate the beauty of Swiss nature, as well as photography.



Selected publications

  1. Recurrent neural networks: vanishing and exploding gradients are not the end of the story
    Nicolas Zucchet, and Antonio Orvieto
    arXiv preprint arXiv:2405.21064, 2024
  2. Gated recurrent neural networks discover attention
    Nicolas Zucchet*Seijin Kobayashi*Yassir Akram*, Johannes Oswald, Maxime Larcher, Angelika Steger, and João Sacramento
    arXiv preprint arXiv:2309.01775, 2023
  3. Online learning of long-range dependencies
    Nicolas Zucchet*Robert Meier*Simon Schug*, Asier Mujika, and João Sacramento
    In Advances in Neural Information Processing Systems , 2023
  4. The least-control principle for local learning at equilibrium
    Alexander Meulemans*Nicolas Zucchet*Seijin Kobayashi*, Johannes Oswald, and João Sacramento
    In Advances in Neural Information Processing Systems , 2022