Constant Bourdrez

I am a PhD student in Machine Learning in the DATA Team at the Centre de Sciences des Données (ENS, PSL University), advised by Olivier Cappé and Alexandre Verine. My research focuses on robustness and privacy in generative models, with an emphasis on image generation. Previously, I studied computer science at the IASD Master and physics at ESPCI Paris-PSL. If you are interested in my research, you can find a complete list of my publications here.

news

Sep 25, 2026	Our paper Learning To Sample From Diffusion Models Via Inverse Reinforcement Learning is accepted at NeurIPS 2026 ! See you in Sydney :)
Feb 10, 2026	New preprint out: Learning To Sample From Diffusion Models Via Inverse Reinforcement Learning, where we proposed a principled IRL framework to optimize diffusion sampling strategies.

selected publications

NeurIPS

Learning To Sample From Diffusion Models Via Inverse Reinforcement Learning

Constant Bourdrez, Alexandre Vérine, and Olivier Cappé

In Advances in Neural Information Processing Systems (NeurIPS), 2026

Bib PDF

@inproceedings{bourdrez2026learningsamplediffusionmodels,
  title = {Learning To Sample From Diffusion Models Via Inverse Reinforcement Learning},
  author = {Bourdrez, Constant and Vérine, Alexandre and Cappé, Olivier},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2026},
  eprint = {2602.08689},
  archiveprefix = {arXiv},
  primaryclass = {cs.LG},
  url = {https://arxiv.org/abs/2602.08689}
}

Workshop

DP-Diffusion Sampling: A Gaussian Proxy Analysis

Constant Bourdrez, Alexandre Vérine, and Olivier Cappé

In InfPriv Workshop @ NeurIPS, 2026

PDF coming soon

Bib

@inproceedings{bourdrez2026dpdiffusion,
  title = {DP-Diffusion Sampling: A Gaussian Proxy Analysis},
  author = {Bourdrez, Constant and Vérine, Alexandre and Cappé, Olivier},
  booktitle = {InfPriv Workshop @ NeurIPS},
  year = {2026},
  note = {PDF coming soon}
}