no code implementations • 8 Feb 2024 • Pierre Marion, Anna Korba, Peter Bartlett, Mathieu Blondel, Valentin De Bortoli, Arnaud Doucet, Felipe Llinares-López, Courtney Paquette, Quentin Berthet
We present a new algorithm to optimize distributions defined implicitly by parameterized stochastic diffusions.
1 code implementation • 3 Sep 2023 • Pierre Marion, Yu-Han Wu, Michael E. Sander, Gérard Biau
Our results are valid for a finite training time, and also as the training time tends to infinity provided that the network satisfies a Polyak-Lojasiewicz condition.
1 code implementation • 19 Apr 2023 • Pierre Marion, Raphaël Berthier
We study the training dynamics of shallow neural networks, in a two-timescale regime in which the stepsizes for the inner layer are much smaller than those for the outer layer.
1 code implementation • 14 Jun 2022 • Pierre Marion, Adeline Fermanian, Gérard Biau, Jean-Philippe Vert
initializations, the only non-trivial dynamics is for $\alpha_L = 1/\sqrt{L}$ (other choices lead either to explosion or to identity mapping).
no code implementations • EMNLP 2021 • Pierre Marion, Paweł Krzysztof Nowak, Francesco Piccinno
On CSQA, our approach increases the coverage from $80\%$ to $96. 2\%$, and the LF execution accuracy from $70. 6\%$ to $75. 6\%$, with respect to previous state-of-the-art results.
1 code implementation • NeurIPS 2021 • Adeline Fermanian, Pierre Marion, Jean-Philippe Vert, Gérard Biau
Building on the interpretation of a recurrent neural network (RNN) as a continuous-time neural differential equation, we show, under appropriate conditions, that the solution of a RNN can be viewed as a linear function of a specific feature set of the input sequence, known as the signature.