Luca Pesce

18 papers A* 5A 1Journal 12
YearRankTypeTitle / Venue / Authors
2026 J jnl
Deep Learning of Compositional Targets with Hierarchical Spectral Methods.
CoRR
Hugo Tabanelli, Yatin Dandi, Luca Pesce, Florent Krzakala
2025 A conf
A Random Matrix Theory Perspective on the Spectrum of Learned Features and Asymptotic Generalization Capabilities.
AISTATS
Yatin Dandi, Luca Pesce, Hugo Cui, Florent Krzakala, Yue M. Lu, Bruno Loureiro
2025 J jnl
The Computational Advantage of Depth: Learning High-Dimensional Hierarchical Functions with Gradient Descent.
CoRR
Yatin Dandi, Luca Pesce, Lenka Zdeborová, Florent Krzakala
2024 J jnl
A Random Matrix Theory Perspective on the Spectrum of Learned Features and Asymptotic Generalization Capabilities.
CoRR
Yatin Dandi, Luca Pesce, Hugo Cui, Florent Krzakala, Yue M. Lu, Bruno Loureiro
2024 A* conf
Asymptotics of feature learning in two-layer networks after one gradient-step.
ICML
Hugo Cui, Luca Pesce, Yatin Dandi, Florent Krzakala, Yue M. Lu, Lenka Zdeborová, Bruno Loureiro
2024 J jnl
Asymptotics of feature learning in two-layer networks after one gradient-step.
CoRR
Hugo Cui, Luca Pesce, Yatin Dandi, Florent Krzakala, Yue M. Lu, Lenka Zdeborová, Bruno Loureiro
2024 J jnl
How Two-Layer Neural Networks Learn, One (Giant) Step at a Time.
J. Mach. Learn. Res.
Yatin Dandi, Florent Krzakala, Bruno Loureiro, Luca Pesce, Ludovic Stephan
2024 J jnl
Online Learning and Information Exponents: On The Importance of Batch size, and Time/Complexity Tradeoffs.
CoRR
Luca Arnaboldi, Yatin Dandi, Florent Krzakala, Bruno Loureiro, Luca Pesce, Ludovic Stephan
2024 A* conf
Online Learning and Information Exponents: The Importance of Batch size & Time/Complexity Tradeoffs.
ICML
Luca Arnaboldi, Yatin Dandi, Florent Krzakala, Bruno Loureiro, Luca Pesce, Ludovic Stephan
2024 J jnl
Repetita Iuvant: Data Repetition Allows SGD to Learn High-Dimensional Multi-Index Functions.
CoRR
Luca Arnaboldi, Yatin Dandi, Florent Krzakala, Luca Pesce, Ludovic Stephan
2024 A* conf
The Benefits of Reusing Batches for Gradient Descent in Two-Layer Networks: Breaking the Curse of Information and Leap Exponents.
ICML
Yatin Dandi, Emanuele Troiani, Luca Arnaboldi, Luca Pesce, Lenka Zdeborová, Florent Krzakala
2024 J jnl
The Benefits of Reusing Batches for Gradient Descent in Two-Layer Networks: Breaking the Curse of Information and Leap Exponents.
CoRR
Yatin Dandi, Emanuele Troiani, Luca Arnaboldi, Luca Pesce, Lenka Zdeborová, Florent Krzakala
2023 A* conf
Are Gaussian Data All You Need? The Extents and Limits of Universality in High-Dimensional Generalized Linear Estimation.
ICML
Luca Pesce, Florent Krzakala, Bruno Loureiro, Ludovic Stephan
2023 J jnl
Are Gaussian data all you need? Extents and limits of universality in high-dimensional generalized linear estimation.
CoRR
Luca Pesce, Florent Krzakala, Bruno Loureiro, Ludovic Stephan
2023 J jnl
Learning Two-Layer Neural Networks, One (Giant) Step at a Time.
CoRR
Yatin Dandi, Florent Krzakala, Bruno Loureiro, Luca Pesce, Ludovic Stephan
2023 J jnl
Theory and applications of the Sum-Of-Squares technique.
CoRR
Francis R. Bach, Elisabetta Cornacchia, Luca Pesce, Giovanni Piccioli
2022 A* conf
Subspace clustering in high-dimensions: Phase transitions & Statistical-to-Computational gap.
NeurIPS
Luca Pesce, Bruno Loureiro, Florent Krzakala, Lenka Zdeborová
2022 J jnl
Subspace clustering in high-dimensions: Phase transitions \& Statistical-to-Computational gap.
CoRR
Luca Pesce, Bruno Loureiro, Florent Krzakala, Lenka Zdeborová