TY  - RPRT
TI  - A three-threshold learning rule approaches the maximal capacity of recurrent neural networks
AU  - Alireza Alemi
AU  - Carlo Baldassi
AU  - Nicolas Brunel
AU  - Riccardo Zecchina
PY  - 2015
DO  - 10.1371/journal.pcbi.1004439
UR  - https://arxiv.org/abs/1508.00429
ID  - 1508.00429
ER  -