TY  - RPRT
TI  - Compositional Function Networks: A High-Performance Alternative to Deep Neural Networks with Built-in Interpretability
AU  - Fang Li
PY  - 2025
UR  - https://arxiv.org/abs/2507.21004
ID  - 2507.21004
ER  -