~~NOCACHE~~

{{page>.:indexheader}}

\\ ==== Prochaines séances ====
[[seminaires:StatP6P7:index|Séminaire de statistique]]\\
Mardi 6 janvier 2026, 10 heures 45, Jussieu en salle 15-16 201\\
**Scott Pesme** (INRIA) //(Old school) deep learning theory through the lens of diagonal linear networks//
<button type="link" size="xs" collapse="b3-175">{{icon>info-circle?lg&color=#777}}</button>\\
<collapse id="b3-175" collapsed="true">
<well size="sm">
Surprisingly, many optimisation phenomena observed in complex neural networks also appear in so-called 2-layer diagonal linear networks. This rudimentary architecture—a two-layer feedforward linear network with a diagonal inner weight matrix—has the advantage of revealing key training characteristics while keeping the theoretical analysis clean and insightful.In this talk, I’ll provide an overview of various theoretical results for this architecture, while drawing connections to experimental observations from practical neural networks. Specifically, we’ll examine how hyperparameters such as the initialisation scale, step size, and batch size impact the optimisation trajectory and influence the generalisation performance of the recovered solution.
</well>
</collapse>


[[seminaires:StatP6P7:index|Séminaire de statistique]]\\
Mardi 20 janvier 2026, 10 heures 45, Sophie Germain en salle 1013\\
**Laurent Oudre** (ENS Paris Saclay) //Non encore annoncé.//
\\


{{page>.:info}}

\\ ==== Séances passées ====
\\ === Année 2025 ===

{{page>.:statp6p72025}}

\\ === Année 2024 ===

{{page>.:statp6p72024}}

\\ === Année 2023 ===

{{page>.:statp6p72023}}

\\ === Année 2022 ===

{{page>.:statp6p72022}}

\\ === Année 2021 ===

{{page>.:statp6p72021}}