Martha White

5

Papers

14

Total Citations

Papers (5)

Scalable Real-Time Recurrent Learning Using Columnar-Constructive Networks

Position: Lifetime tuning is incompatible with continual reinforcement learning

Averaging $n$-step Returns Reduces Variance in Reinforcement Learning

Position: Benchmarking is Limited in Reinforcement Learning Research

Scalable Real-Time Recurrent Learning Using Columnar-Constructive Networks