Research summary

Guillermo's research studies communication and cooperation in multi-agent systems under uncertainty and strategic incentives. He investigates when communication supports coordination, when it becomes misleading or exploitable, and how to design mechanisms that make agent interactions more reliable.

Bio

Guillermo is a PhD candidate at ENS Paris-Saclay and the International Laboratory on Learning Systems, associated with Mila - Quebec AI Institute and ÉTS Montréal. Before starting his PhD, he completed an MSc in Mathematical Sciences at the University of Oxford.

Bio

Tristan Leuridan is a PhD candidate at the Laboratoire Génie Industriel of CentraleSupélec, supervised by Yiping Fang and Adam Abdin. He holds a Bachelor's in Mechanical Engineering from the University of Southampton and an MSc in Aerospace Engineering from Imperial College London. Before his PhD, he worked as a ground-station operator at the DLR (German Aerospace Center).

Research summary

Ali's PhD develops AI models that generate high-quality 3D human animation from text prompts, making animation production accessible beyond studios with motion-capture equipment. His approach uses VAE-based, diffusion-based and hybrid methods with 3D Gaussian Splatting as the motion representation.

Bio

Ali Rida Sahili is a PhD candidate from Lebanon at the IBISC Laboratory, University of Évry Paris-Saclay. His doctoral research is co-funded by Magic Factory, a startup specialising in metaverse technologies. He previously earned a Master's from ENS Paris-Saclay in Mathematics, Machine Learning, and Computer Vision.

Research summary

Quoc-Duong's PhD develops climate downscaling methods using generative models, transforming coarse global climate model outputs into fine-scale local projections for agriculture, urban planning and disaster management. His approach combines model-based statistical methods with modern machine learning, improving predictive performance while keeping models interpretable.

Bio

Quoc-Duong Nguyen is a PhD candidate from Vietnam at L2S, supervised by Mohammed Nabil El Korso and colleagues. He received his Master's in Applied Data Science from Quy Nhon University, graduating as valedictorian, and was selected for the France Excellence Eiffel Scholarship in 2025.

Research summary

Adil's PhD develops accelerated diffusion processes for uncertainty quantification in image reconstruction, using low-rank models and Bayesian frameworks to tackle ill-posed inverse problems such as those arising in MRI and interferometry.

Bio

Adil Zomahoun is a PhD student from Benin at L2S, supervised by François Orieux and Jérôme Idier. He graduated from the École Centrale Casablanca and has prior experience in data analysis, LLM-based legal search automation, and applied AI research at CEA-ISEC.

Research summary

Lorenzo's PhD develops neural network methods for learning causal relationships in complex, nonlinear, non-Markovian systems, using autoencoders to find coordinate transformations where nonlinear stochastic processes become linear in latent space.

Bio

Lorenzo Mensi is a PhD candidate from Italy at LISN, supervised by Sergio Chibbaro and Cyril Furtlerhner. He trained as a physicist at the University of Bologna, then at Politecnico di Torino and Université Paris-Saclay in Physics of Complex Systems.

Research summary

Raffael's PhD targets object detection with design-explainable models, focusing on Concept Bottleneck Models (CBMs). He introduced the Irrelevant Concept Contribution (ICC) metric to detect and quantify information leakage, preserving accuracy while reducing it.

Bio

Raffael Schoen is a PhD candidate from Germany at ONERA, supervised by Stéphane Herbin and Baptiste Abeloos. He trained in Business Administration and Business Informatics, with prior experience in software engineering at Deloitte, EY, NLP at Continental AG, and AI engineering.

Bio

Jinu Raj is a PhD candidate from India at CEA Paris-Saclay (DPhP), supervised by Frédéric Deliot in the ATLAS group. He completed an Integrated Master of Science in Physics at the Central University of Tamil Nadu. He has held research positions at INFN Bologna and INFN Padova, working on unsupervised learning for particle-trajectory identification.

Research summary

Pietro's PhD uses machine learning to study the Epoch of Reionisation, developing inference frameworks that exploit full simulated 21cm-signal maps rather than summary statistics, in preparation for the Square Kilometre Array data.

Bio

Pietro Albanese Guidi trained at the University of Groningen (BSc Physics) and ETH Zürich (MSc Astrophysics), with research internships at ENS Paris and the Max Planck Institute for Astrophysics. He is now a PhD candidate at the Institut d'Astrophysique Spatiale.

Research summary

Quân's PhD builds an interpretable, real-time digital twin of a patient's heart, coupling the modified Constitutive Relation Error concept with Kolmogorov-Arnold networks inside an Isogeometric Analysis framework to support personalised cardiac care.

Bio

Hoàng Quân Pham is a PhD candidate from Vietnam at LMPS, supervised by Cuong Ha-Minh. He holds an EUR-ACE Master's from the Hanoi University of Science and Technology, with an Erasmus+ exchange to the University of Ruse in Bulgaria.

Research summary

Marouane's PhD focuses on detecting out-of-distribution data in multi-label classification systems for non-destructive artwork analysis. He introduced HiPOOD, a zero-shot hierarchical detector accepted at IEEE ICIP 2026.

Bio

Marouane Hadj-Ali is a PhD candidate from Algeria at SATIE, supervised by Florence Alberge. He earned a Master's in Data Science at Université Paris-Saclay, with prior experience in computer vision, NLP, and image super-resolution.

Research summary

Constanza's PhD develops physics-informed reduced-order models for digital twins of sustainable combustion systems, combining uncertainty quantification and data assimilation to predict and explain flame transitions in aeronautical combustors.

Bio

Constanza Lopez is a PhD candidate from Chile at EM2C, supervised by Bérengère Podvin and Salvatore Iavarone. She holds a Bachelor's and Master's in Industrial Engineering from Universidad Técnica Federico Santa María.

Research summary

Jiarong's research focuses on conformal prediction for machine learning models handling missing data. He developed a mask-conditional method accepted at AISTATS 2026, with applications to streamflow forecasting at monitored and ungauged river stations.

Bio

Jiarong Fan is a PhD candidate from China at LaMME. He holds a BSc in Mathematics from Beihang University, a double Master's in Engineering, and an M2 in Probability and Finance from Sorbonne University. He previously worked as a data scientist at China Merchants Bank.

Research summary

Harlette's PhD develops methods to integrate trustworthy AI into clinical decision-making for patient profiling from heterogeneous healthcare data such as genomics and imaging, focusing on multimodal learning, interpretability and robustness.

Bio

Harlette Denebeye is a PhD candidate from Chad at CEA-CNRGH. She holds a BSc in Computer Science from the University of Yaoundé I and was awarded a Google DeepMind Scholarship for her Master's in AI for Science at AIMS, South Africa.

Research summary

Minh Lien's PhD develops physics-informed machine learning models, including PINNs, to accelerate dynamic power-system simulations while maintaining accuracy, addressing the growing complexity brought by renewable energy sources.

Bio

Thi Minh Lien Le is a PhD candidate from Vietnam at GeePs, supervised by Philippe Dessante and Trung-Dung Le. She graduated in the top 5% of her department from HUST and received the Best Thesis Award in two consecutive years (2023 and 2024).

Research summary

Yvon's PhD develops lightweight Vision-Language Models (VLMs) and Vision-Language-Action (VLA) models for resource-constrained embedded devices, combining compression, efficient fine-tuning and runtime optimisations.

Bio

Yvon Apedo is a PhD candidate from Togo at IBISC and CEA LIAE. He holds a Master's in Computer Science from Northwestern Polytechnical University. His prior research covered Transformers, weakly-supervised crack segmentation and unsupervised domain adaptation.

Research summary

Yuliia's PhD develops a transformer-based photon reconstruction algorithm for the CMS detector at the LHC, substantially improving upon the traditional algorithm in complex topologies while integrating high-precision timing from the new MTD detector.

Bio

Yuliia Maidannyk is a PhD candidate at CEA Paris-Saclay (DPhP), supervised by Fabrice Couderc and Özgür Sahin in the CMS group. She holds an MSc in Physics from ETH Zurich and a BSc in Physics with Theoretical Physics from the University of Manchester.

Research summary

Sammy's PhD develops generative AI methods for robust uncertainty quantification in astrophysical inverse problems, focusing on strong gravitational lensing and building a full Bayesian framework combining physics-constrained likelihoods with generative priors.

Bio

Sammy Sharief is a PhD candidate at AIM, supervised by François Lanusse, Tobias Liaudat, and Samuel Farrens. He holds a BSc in Computer Science and Astronomy from UIUC and a Master's in Computer Science from Université de Montréal. He is also affiliated with MILA.

Research summary

Nived's PhD develops the ML reconstruction pipeline for the DUNE liquid-argon time-projection chamber, using U-ResNet segmentation, point-proposal networks and graph neural networks to turn raw 3D detector hits into labelled particles.

Bio

Nived Puthumana Meleppattu is a PhD candidate from India at IJCLab, supervised by Yoann Kermaïdic in the DUNE group. He holds a Master's in High-Energy Physics from ETH Zurich and IP Paris, with prior research on T2K and CMS experiments.

Research summary

Laurent's PhD develops explainable deep learning methods for dynamic MRI reconstruction, exploring physics-informed neural networks and neural operators to accelerate acquisition while maintaining diagnostic image quality and clinician trust.

Bio

Laurent Brock is a life-sciences engineer from EPFL (Bachelor's and Master's). His Master's thesis was carried out at Harvard Medical School's Nezami Lab. Earlier projects included clinical data processing software and batch-effect correction in single-cell sequencing at EPFL's Courtine Lab.