TY  - RPRT
TI  - Implicit bias produces neural scaling laws in learning curves, from perceptrons to deep networks
AU  - Francesco D'Amico
AU  - Dario Bocchi
AU  - Matteo Negri
PY  - 2026
UR  - https://arxiv.org/abs/2505.13230
ID  - 2505.13230
ER  -