Welcome to the Operational Research and AI Lab at LIX
ORAILIX is a newly established AI research team at the Laboratoire d’Informatique de l’X (LIX) at École Polytechnique, directed by Prof. Sonia Vanier. Founded in September 2024, our research projects aim to develop new hybrid approaches that combine the power of machine learning algorithms and architectures with the robustness of operational research methods.
News
Our mission is to develop secure, ethical, and robust AI-based systems that address real-world challenges.
AI Summit 2025
ORAILIX will be present at the summit 6-7 Feb at Ecole Polytechnique.
ORAILIX Presents Hybrid AI-OR Approaches at Google
We highlighted a pipeline we developed to identify and address memory-related issues in large language models.
Publications
- I-GLIDE: Input Groups for Latent Health Indicators in Degradation Estimation accepted at ECML PKDD 2025 for presentation.
- Uncertainty Quantification as a Complementary Latent Health Indicator for Remaining Useful Life Prediction on Turbofan Engines has been accepted at MIMAR 2025.
- PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models" has been accepted at CVPR 2025.
- Dentan, Jérémie, et al. "Predicting and analyzing memorization within fine-tuned Large Language Models." arXiv preprint arXiv:2409.18858 (2024).
- Dentan, Jérémie, Arnaud Paran, and Aymen Shabou. "Reconstructing training data from document understanding models." 33rd USENIX Security Symposium (USENIX Security 24). 2024.
Our two Research Chairs
Press
