Unsupervised Predictive Memory in a Goal-Directed Agent

by   Greg Wayne, et al.

Animals execute goal-directed behaviours despite the limited range and scope of their sensors. To cope, they explore environments and store memories maintaining estimates of important information that is not presently available. Recently, progress has been made with artificial intelligence (AI) agents that learn to perform tasks from sensory input, even at a human level, by merging reinforcement learning (RL) algorithms with deep neural networks, and the excitement surrounding these results has led to the pursuit of related ideas as explanations of non-human animal learning. However, we demonstrate that contemporary RL algorithms struggle to solve simple tasks when enough information is concealed from the sensors of the agent, a property called "partial observability". An obvious requirement for handling partially observed tasks is access to extensive memory, but we show memory is not enough; it is critical that the right information be stored in the right format. We develop a model, the Memory, RL, and Inference Network (MERLIN), in which memory formation is guided by a process of predictive modeling. MERLIN facilitates the solution of tasks in 3D virtual reality environments for which partial observability is severe and memories must be maintained over long durations. Our model demonstrates a single learning agent architecture that can solve canonical behavioural tasks in psychology and neurobiology without strong simplifying assumptions about the dimensionality of sensory input or the duration of experiences.


page 11

page 14


Reinforcement Learning using Guided Observability

Due to recent breakthroughs, reinforcement learning (RL) has demonstrate...

Universal Memory Architectures for Autonomous Machines

We propose a self-organizing memory architecture for perceptual experien...

Zipfian environments for Reinforcement Learning

As humans and animals learn in the natural world, they encounter distrib...

Episodic Memory Deep Q-Networks

Reinforcement learning (RL) algorithms have made huge progress in recent...

POPGym: Benchmarking Partially Observable Reinforcement Learning

Real world applications of Reinforcement Learning (RL) are often partial...

Learning to Navigate in Complex Environments

Learning to navigate in complex environments with dynamic elements is an...

Please sign up or login with your details

Forgot password? Click here to reset