Our AI Research Lab
Our R&D Lab is dedicated to research and innovation on the most advanced artificial intelligence technologies.
ICML, NeurIPS, ECML, PKDD
AI Lab: Science, Innovation, Solutions
For the past 10 years, Craft AI has been pushing the boundaries of AI research through its dedicated R&D lab. Our engineers regularly publish in top Machine Learning conferences, with a focus on trustworthy AI that emphasizes privacy, causality, explainability, and efficiency.
A pioneering team in trustworthy AI
Our R&D Lab drives the exploration and development of cutting-edge artificial intelligence technologies.
Explainability
Ensuring AI system decisions and predictions are understandable and interpretable.
Privacy
Safeguarding personal and sensitive data in AI systems to guarantee user security and privacy.
Robustness
The ability of an artificial intelligence system to function correctly and withstand errors, disruptions, or unexpected attacks.
Our most recent research publications
Explore our team’s research and scientific publications from recent years.
2026
A Divide-and-Conquer Framework for Query-Efficient Causal Discovery with Large Language Models - AIDT
Louis Hernandez, Matthieu Boussard, Alessandro Leite
2026
Privacy Amplification Persists under Unlimited Synthetic Data Release - arXiv
Clément Pierquin, Aurélien Bellet, Marc Tommasi, Matthieu Boussard
2025
Needle in a Patched Haystack: Evaluating Saliency Maps for Vision LLMs - AAAI/ACM AIES
Bastien Zimmermann, Matthieu Boussard
2025
Privacy Amplification Through Synthetic Data: Insights from Linear Regression - ICML
Clément Pierquin, Aurélien Bellet, Marc Tommasi, Matthieu Boussard
2025
Transplanting Knowledge: A Study on Layer-Specific Grafting in LLMs
Bastien Zimmermann, Matthieu Boussard
2024
The impact of lora on the emergence of clusters in transformers - arXiv
Hugo Koubbi, Matthieu Boussard, Louis Hernandez
2024
Rényi pufferfish privacy: General additive noise mechanisms and privacy amplification by iteration via shift reduction lemmas - ICML
Clément Pierquin, Aurélien Bellet, Marc Tommasi, Matthieu Boussard
2023
CausalStructCodec: Causally-aware observational and interventional data generator - ICLR
Louis Hernandez, Matthieu Boussard
2022
Practical considerations on using private sampling for synthetic data - ICML
Clément Pierquin, Bastien Zimmermann, Matthieu Boussard
2022
XAI and geographic information: application to paleoenvironmental reconstructions - ECML/PKDD
Bastien Zimmermann, Matthieu Boussard, Nicolas Boulbes, Sophie Grégoire
2022
Improving drift detection by monitoring shapley loss values - International Conference on Pattern Recognition and Artificial Intelligence
Bastien Zimmermann, Matthieu Boussard
