CB-RNNs with a cerebellar feedforward module learn temporal tasks faster than matched RNNs, with the module driving efficiency even after freezing the recurrent core as a fixed reservoir.
Cerebro- cerebellar networks facilitate learning through feedback decoupling.Nature Communications, 14(1):51, January 2023
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
q-bio.NC 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Cortico-cerebellar modularity as an architectural inductive bias for efficient temporal learning
CB-RNNs with a cerebellar feedforward module learn temporal tasks faster than matched RNNs, with the module driving efficiency even after freezing the recurrent core as a fixed reservoir.