Jacopo Iollo
Personal page
Very Short Bio
I am a PhD student (graduated in May) @INRIA Grenoble-Alpes in the STATIFY team. Prior to that, I studied Maths & Computer Science at Grenoble INP - Ensimag and Applied Mathematics at Imperial College London.
Living in Grenoble, France, I can be found skiing during winter and cycling during summer.
You can reach me at: jacopo.iollo [at] inria.fr
Research Interests
My research focuses on computational methods for Bayesian inference, with a particular emphasis on Bayesian Experimental Design (BED). I am interested in how to design efficient experiments or measurements to maximize information gain under limited data and computational budgets.
This naturally leads me to questions at the interface of:
- Sampling and optimization methods and their connections
- Diffusion model sampling for inverse problems and scientific imaging
- Efficient diffusion architectures and training strategies, especially under conditioning or structured constraints
- Generative modeling approaches for scientific computation and uncertainty quantification
I am also interested in the practical deployment of BED and diffusion-based methods in real experimental systems, particularly in imaging and engineering applications.
News
2025
-
Nov 2025 – 📘 New tutorial released: “Quick experimentation with Flux using diffuse” showing how to use the diffuse library to rapidly prototype with the Flux text-to-image model.
-
Jul 2025 – 🗣️ Invited talk “Computational Methods for Bayesian Experimental Design” at Monte Carlo Methods Conference 2025, Chicago, USA · 📄 Preprint: Active MRI Acquisition with Diffusion Guided Bayesian Experimental Design (with G. Oudoumanessah et al.)
-
Jun 2025 – 🎙️ Invited keynote “Diffusion Based Bayesian Experimental Design” at the Isaac Newton Institute for Mathematical Sciences, Cambridge, UK · 🎥 Watch on YouTube · 📢 Invited talk “Advancing Bayesian Experimental Design for Practical Applications” at IMS Spring Research Conference 2025, New York, USA
-
May 2025 – 🎓 PhD defended at Inria / Université Grenoble-Alpes: “Inference driven Bayesian Experimental Design”
-
Apr 2025 – 🧠 ICLR 2025 Spotlight (top 5%) for “Bayesian Experimental Design via Contrastive Diffusions”
-
Jan 2025 – 📢 Talk “Bi-Level Optimization Meets Diffusions for Tractable Bayesian Experimental Design” at Mostly Monte Carlo Seminar, Université Paris Dauphine
2024
- Sep–Nov 2024 – 🎓 Visiting scholar at Carnegie Mellon University
- Jul 2024 – 📄 PASOA: Particle-Based Bayesian Optimal Adaptive Design, ICML 2024, Vienna (poster)
- Jun 2024 – 🧪 Talk “PASOA: Particle-Based Bayesian Optimal Adaptive Design” at 2nd Bayes-Duality Workshop, Tokyo
Publications
- Active MRI Acquisition with Diffusion-Guided Bayesian Experimental Design, Jacopo Iollo, G. Oudoumanessah, et al., Preprint 2025
- Bayesian Experimental Design via Contrastive Diffusions, Jacopo Iollo, Christophe Heinkelé, Pierre Alliez, Florence Forbes, ICLR 2025 Spotlight
- PASOA: PArticle baSed Bayesian Optimal Adaptive design, Jacopo Iollo, Christophe Heinkelé, Pierre Alliez, Florence Forbes ICML 2024
- Tempered SMC for Sequential Bayesian Optimal Design, Jacopo Iollo, Christophe Heinkelé, Pierre Alliez, Florence Forbes GRETSI 2023