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

Leveraging learned strategies in unfamiliar scenarios is fundamental to humanintelligence. In reinforcement learning, rationally reusing the policiesacquired from other tasks or human experts is critical for tackling problemsthat are difficult to learn from scratch. In this work, we present a frameworkcalled Selective Myopic bEhavior Control~(SMEC), which results from the insightthat the short-term behaviors of prior policies are sharable across tasks. Byevaluating the behaviors of prior policies via a hybrid value functionarchitecture, SMEC adaptively aggregates the sharable short-term behaviors ofprior policies and the long-term behaviors of the task policy, leading tocoordinated decisions. Empirical results on a collection of manipulation andlocomotion tasks demonstrate that SMEC outperforms existing methods, andvalidate the ability of SMEC to leverage related prior policies.

On the Value of Myopic Behavior in Policy Reuse