A Unified Bellman Optimality Principle Combining Reward Maximization and Empowerment

Abstract

Empowerment is an information-theoretic method that can be used to intrinsically motivate learning agents. It attempts to maximize an agent's control over the environment by encouraging visiting states with a large number of reachable next states. Empowered learning has been shown to lead to complex behaviors, without requiring an explicit reward signal. In this paper, we investigate the use of empowerment in the presence of an extrinsic reward signal. We hypothesize that empowerment can guide reinforcement learning (RL) agents to find good early behavioral solutions by encouraging highly empowered states. We propose a unified Bellman optimality principle for empowered reward maximization. Our empowered reward maximization approach generalizes both Bellman's optimality principle as well as recent information-theoretical extensions to it. We prove uniqueness of the empowered values and show convergence to the optimal solution. We then apply this idea to develop off-policy actor-critic RL algorithms which we validate in high-dimensional continuous robotics domains (MuJoCo). Our methods demonstrate improved initial and competitive final performance compared to model-free state-of-the-art techniques.

arXiv commentProceedings of the 33rd Conference on Neural Information Processing Systems (NeurIPS), Vancouver, Canada, 2019

References in corpus (20)

• Auto-Encoding Variational Bayes
• Continuous control with deep reinforcement learning
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• Trust Region Policy Optimization
• Addressing Function Approximation Error in Actor-Critic Methods
• Soft Actor-Critic Algorithms and Applications
• Reinforcement Learning with Deep Energy-Based Policies
• Reinforcement Learning and Control as Probabilistic Inference: Tutorial and Review
• Variational Information Maximisation for Intrinsically Motivated Reinforcement Learning
• Maximum a Posteriori Policy Optimisation
• Bridging the Gap Between Value and Policy Based Reinforcement Learning
• Equivalence Between Policy Gradients and Soft Q-Learning
• A unified view of entropy-regularized Markov decision processes
• Off-Policy Actor-Critic
• A unified strategy for implementing curiosity and empowerment driven reinforcement learning
• A Unified Bellman Equation for Causal Information and Value in Markov Decision Processes
• An Information-Theoretic Optimality Principle for Deep Reinforcement Learning
• Information Theoretically Aided Reinforcement Learning for Embodied Agents
• Bounded Rational Decision-Making in Feedforward Neural Networks
• Empowerment-driven Exploration using Mutual Information Estimation

Cited by in corpus (4)

• Action and Perception as Divergence Minimization
• A Tutorial on Sparse Gaussian Processes and Variational Inference
• Decoupling Exploration and Exploitation for Meta-Reinforcement Learning without Sacrifices
• Policy Information Capacity: Information-Theoretic Measure for Task Complexity in Deep Reinforcement Learning