Scholar
Umut Simsekli
Google Scholar ID: CuArAkgAAAAJ
INRIA - École Normale Supérieure
Deep Learning Theory
Langevin Monte Carlo
Follow
Homepage
↗
Google Scholar
↗
Citations & Impact
All-time
Citations
3,297
H-index
32
i10-index
53
Publications
20
Co-authors
18
list available
Contact
No contact links provided.
Publications
9 items
Algorithm- and Data-Dependent Generalization Bounds for Score-Based Generative Models
2025
Cited
0
Understanding the Generalization Error of Markov algorithms through Poissonization
2025
Cited
0
Algorithmic Stability of Stochastic Gradient Descent with Momentum under Heavy-Tailed Noise
2025
Cited
0
The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training
2025
Cited
0
Denoising Lévy Probabilistic Models
arXiv.org · 2024
Cited
0
Uniform Generalization Bounds on Data-Dependent Hypothesis Sets via PAC-Bayesian Theory on Random Sets
arXiv.org · 2024
Cited
0
Privacy of SGD under Gaussian or Heavy-Tailed Noise: Guarantees without Gradient Clipping
2024
Cited
2
A PAC-Bayesian Link Between Generalisation and Flat Minima
arXiv.org · 2024
Cited
3
Load more
Resume (English only)
Co-authors
18 total
Co-author 1
Gaël Richard
Professor, Télécom Paris, Institut polytechnique de Paris
Mert Gürbüzbalaban
Rutgers University
Alain Durmus
Ecole polytechnique
Lingjiong Zhu
Florida State University
Roland Badeau
Full Professor of Signal Processing, Telecom Paris
Kimia Nadjahi
CNRS - ENS Paris
Co-author 8
×
Welcome back
Sign in to Agora
Welcome back! Please sign in to continue.
Email address
Password
Forgot password?
Continue
Do not have an account?
Sign up