The Dormant Neuron Phenomenon in Deep Reinforcement Learning.
Ghada SokarRishabh AgarwalPablo Samuel CastroUtku EvciPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- neural network
- function approximation
- reinforcement learning algorithms
- markov decision processes
- state space
- optimal control
- robotic control
- multi agent
- model free
- neural model
- policy search
- learning algorithm
- optimal policy
- learning agents
- dynamic programming
- learning problems
- search space
- action selection
- temporal difference
- case study
- deep learning
- reinforcement learning methods
- stochastic approximation