English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In reinforcement learning (RL), the ability to utilize prior knowledge frompreviously solved tasks can allow agents to quickly solve new problems. In somecases, these new problems may be approximately solved by composing thesolutions of previously solved primitive tasks (task composition). Otherwise,prior knowledge can be used to adjust the reward function for a new problem, ina way that leaves the optimal policy unchanged but enables quicker learning(reward shaping). In this work, we develop a general framework for rewardshaping and task composition in entropy-regularized RL. To do so, we derive anexact relation connecting the optimal soft value functions for twoentropy-regularized RL problems with different reward functions and dynamics.We show how the derived relation leads to a general result for reward shapingin entropy-regularized RL. We then generalize this approach to derive an exactrelation connecting optimal value functions for the composition of multipletasks in entropy-regularized RL. We validate these theoretical contributionswith experiments showing that reward shaping and task composition lead tofaster learning in various settings.

Utilizing Prior Solutions for Reward Shaping and Composition in Entropy-Regularized Reinforcement Learning