DeepAI AI Chat
Log In Sign Up

Understanding the impact of entropy on policy optimization

by   Zafarali Ahmed, et al.

Entropy regularization is commonly used to improve policy optimization in reinforcement learning. It is believed to help with exploration by encouraging the selection of more stochastic policies. In this work, we analyze this claim and, through new visualizations of the optimization landscape, we observe that incorporating entropy in policy optimization serves as a regularizer. We show that even with access to the exact gradient, policy optimization is difficult due to the geometry of the objective function. We qualitatively show that, in some environments, entropy regularization can make the optimization landscape smoother, thereby connecting local optima and enabling the use of larger learning rates. This manuscript presents new tools for understanding the underlying optimization landscape and highlights the challenge of designing general-purpose policy optimization algorithms in reinforcement learning.


page 1

page 2

page 3

page 4


Understanding the impact of entropy in policy learning

Entropy regularization is commonly used to improve policy optimization i...

EnTRPO: Trust Region Policy Optimization Method with Entropy Regularization

Trust Region Policy Optimization (TRPO) is a popular and empirically suc...

Implicit Policy for Reinforcement Learning

We introduce Implicit Policy, a general class of expressive policies tha...

Marginalized State Distribution Entropy Regularization in Policy Optimization

Entropy regularization is used to get improved optimization performance ...

Regularization Matters in Policy Optimization

Deep Reinforcement Learning (Deep RL) has been receiving increasingly mo...

An Entropy Regularization Free Mechanism for Policy-based Reinforcement Learning

Policy-based reinforcement learning methods suffer from the policy colla...

Taylor Expansion Policy Optimization

In this work, we investigate the application of Taylor expansions in rei...