Publications
- Distillation of Weighted Automata from Recurrent Neural Networks using a Spectral Approach, Rémi Eyraud, Stéphane Ayache, Machine Learning Journal, 2021
- Connecting Weighted Automata, Tensor Networks and Recurrent Neural Networks through Spectral Learning, Tianyu Li, Doina Precup and Guillaume Rabusseau. Machine Learning Journal, 2021
- Explicabilité dans les réseaux récurrents par discrétisation, Hamed Benazha, Stéphane Ayache, Rémi Eyraud and Thierry Artières, Cap 2022
- Sur les limites de la descente de gradient en précision finie pour l'apprentissage de réseaux récurrents, Rémi Eyraud and Volodimir Mitarchuk, Cap 2022
- On the limit of gradient descent for Simple Recurrent Neural Networks with finite precision, Rémi Eyraud and Volodimir Mitarchuk, LearnAut 2022
- Spectral Regularization: an Inductive Bias for Sequence Modeling, Kaiwen Hou and Guillaume Rabusseau, LearnAut 2022
- Spectral Initialization of Recurrent Neural Networks: Proof of Concept, Maude Lizaire, Simon Verret and Guillaume Rabusseau. LearnAut 2022
- Towards an AAK Theory Approach to Approximate Minimization in the Multi-Letter Case, Clara Lacroce, Prakash Panangaden and Guillaume Rabusseau, LearnAut 2022
- Calibrate to Interpret, Gregory Scafarto, Nicolas Posocco, Antoine Bonnefoy, European Conference on Machine Learning, 2022