Noisy Nets

E1849526 UNEXPLORED

Noisy Nets is a neural network technique that injects learnable, parameterized noise into network weights to enable more efficient and exploratory reinforcement learning.

Try in SPARQL Jump to: Surface forms Referenced by

All labels observed (1)

Label	Occurrences
Noisy Nets canonical	1

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.

Rainbow DQN → combinesMethod → Noisy Nets ⓘ

subject linked to: Dan Horgan