Assistant Professor, University of Venice
1 paper at NeurIPS 2025
Equivariant neural networks can fail to approximate target functions even when they achieve maximal separation power.
