TY  - RPRT
TI  - Thermodynamic efficiency of learning a rule in neural networks
AU  - Sebastian Goldt
AU  - Udo Seifert
PY  - 2017
DO  - 10.1088/1367-2630/aa89ff
UR  - https://arxiv.org/abs/1706.09713
ID  - 1706.09713
ER  -