Thèse, AO, Mathematics, Optimization and Control

Spotting expressivity bottlenecks in neural networks and fixing them by optimal architecture growth

Manon Verbockhaven. Spotting expressivity bottlenecks in neural networks and fixing them by optimal architecture growth. Neural and Evolutionary Computing [cs.NE]. Université Paris-Saclay, 2025. English. ⟨NNT : 2025UPASG022⟩. ⟨tel-05232707⟩

Published on