4 papers from Maasai team at ICML!

Published on May 26, 2023 Updated on May 30, 2023

3IA Chair papers accepted at the Fortieth International Conference on Machine Learning.

The Inria Maasai team had 4 papers accepted at ICML 2023, two of them involve 3IA-funded researchers! Topics include interpretability (Senetaire et al., Barbiero et al., 2023), NLP (Catellier et al., 2023), and semi-supervised learning (Sportisse et al., 2023). The conference will be held on July 23 in Hawaii.

The papers: 
Barbiero et al., Interpretable Neural-Symbolic Concept Reasoning , ICML 2023, https://arxiv.org/abs/2304.14068
Catellier et al., On the Robustness of Text Vectorizers, ICML 2023, https://arxiv.org/abs/2303.07203
Senetaire et al., Explainability as statistical inference, ICML 2023, https://arxiv.org/abs/2212.03131
Sportisse et al., Are labels informative in semi-supervised learning? Estimating and leveraging the missing-data mechanism, ICML 2023, https://arxiv.org/abs/2302.07540