Assistant Professor, Télécom Paris
1 paper at NeurIPS 2025
A cognition grounded interpretability method