A single evolved Hebbian update rule, shared by all agents, produces heterogeneous neural controllers that outperform homogeneous evolution and MARL baselines in simulated swarm tasks and in a small real-robot test.
Nature communications 12(1), 4345 (2021)
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
cs.NE 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Emergent Heterogeneous Swarm Control Through Hebbian Learning
A single evolved Hebbian update rule, shared by all agents, produces heterogeneous neural controllers that outperform homogeneous evolution and MARL baselines in simulated swarm tasks and in a small real-robot test.