Options of Interest: Temporal Abstraction with Interest Functions
Posted on10 Jan 2020
Temporal abstraction refers to the ability of an agent to use behaviours of controllers which act for a limited, variable amount of...Read More
Leveraging exploration in off-policy algorithms via normalizing flows
Posted on24 Oct 2019
The ability to discover approximately optimal policies in domains with sparse rewards is crucial to applying reinforcement learning (RL) in many realworld...Read More
Learning Causal State Representations of Partially Observable Environments
Posted on25 Jun 2019
Intelligent agents can cope with sensory-rich environments by learning task-agnostic state abstractions. In this paper, we propose mechanisms to approximate causal states,...Read More
Learning Powerful Policies by Using Consistent Dynamics Mode
Posted on11 Jun 2019
Model-based Reinforcement Learning approaches have the promise of being sample efficient. Much of the progress in learning dynamics models in RL has...Read More
Variational State Encoding as Intrinsic Motivation in Reinforcement Learning
Posted on06 May 2019
Discovering efficient exploration strategies is a central challenge in reinforcement learning (RL), especially in the context of sparse rewards environments. We postulate...Read More
Active Domain Randomization
Posted on09 Apr 2019
Domain randomization is a popular technique for improving domain transfer, often used in a zero-shot setting when the target domain is unknown...Read More
Hyperbolic Discounting and Learning over Multiple Horizons
Posted on19 Feb 2019
Reinforcement learning (RL) typically defines a discount factor as part of the Markov Decision Process. The discount factor values future rewards by...Read More
Separating value functions across time-scales
Posted on05 Feb 2019
In many finite horizon episodic reinforcement learning (RL) settings, it is desirable to optimize for the undiscounted return – in settings like...Read More
Prioritizing Starting States for Reinforcement Learning
Posted on27 Nov 2018
Online, off-policy reinforcement learning algorithms are able to use an experience memory to remember and replay past experiences. In prior work, this...Read More
Count-Based Exploration with the Successor Representation
Posted on31 Jul 2018
In this paper we introduce a simple approach for exploration in reinforcement learning (RL) that allows us to develop theoretically justified algorithms...Read More