To interact with humans in the world, agents need to understand the dive...
Specifying reward signals that allow agents to learn complex behaviors i...
General intelligence requires solving tasks across many domains. Current...
Intelligent agents need to remember salient information to reason in
par...
Generating long, temporally consistent video remains an open challenge i...
Visual model-based reinforcement learning (RL) has the potential to enab...
To solve tasks in complex environments, robots need to learn from experi...
Intelligent agents need to select long sequences of actions to solve com...
Humans and animals explore their environment and acquire useful skills e...
How can artificial agents learn to solve many diverse tasks in complex v...
Evaluating the general abilities of intelligent agents requires complex
...
Deep learning has enabled algorithms to generate realistic images. Howev...
Reinforcement learning has enabled agents to solve challenging tasks in
...
Model-based reinforcement learning (MBRL) methods have shown strong samp...
To quickly solve new tasks in complex environments, intelligent agents n...
Intelligent agents need to generalize from past experience to achieve go...
We introduce a unified objective for action and perception of intelligen...
Active inference offers a first principle account of sentient behaviour,...
Reinforcement learning allows solving complex tasks, however, the learni...
Learned world models summarize an agent's experience to facilitate learn...
We describe Bayesian Layers, a module designed for fast experimentation ...
Solving tasks with sparse rewards is a main challenge in reinforcement
l...
Planning has been very successful for control tasks with known environme...
Obtaining reliable uncertainty estimates of neural network predictions i...
Integrating model-free and model-based approaches in reinforcement learn...
Designing agile locomotion for quadruped robots often requires extensive...
Learning distributed representations of documents has pushed the
state-o...
We introduce TensorFlow Agents, an efficient infrastructure paradigm for...
We propose ThalNet, a deep learning model inspired by neocortical
commun...
Using current reinforcement learning methods, it has recently become pos...