Demis Hassabis

deep reinforcement learning (2) sample efficiency (1) speech synthesis (1) value function (1) episodic memory (1) state representation (1) model-based reinforcement learning (1) logarithmic loss (1) model-free reinforcement learning (1) feed-forward network (1) task identification (1) experience replay (1) bayesian online learning (1) waveform generation (1) model-based rl (1) neural episodic control (1) semi-tabular representation (1) environment model (1) parallel neural network (1) probabilistic sequence prediction (1)