Jakob N. Foerster

reinforcement learning (2) offline reinforcement learning (1) linear methods (1) model-based reinforcement learning (1) value function approximation (1) value estimation (1) text modeling (1) recurrent neural network (1) learned optimization (1) update rule (1) plasticity loss (1) world model (1) value overestimation (1) edge-of-reach problem (1) dynamics model (1) target network (1) learned optimizer (1) neural network optimizer (1) affine transformation (1) gradient magnitude (1)