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

The exploration problem is one of the main challenges in deep reinforcementlearning (RL). Recent promising works tried to handle the problem withpopulation-based methods, which collect samples with diverse behaviors derivedfrom a population of different exploratory policies. Adaptive policy selectionhas been adopted for behavior control. However, the behavior selection space islargely limited by the predefined policy population, which further limitsbehavior diversity. In this paper, we propose a general framework calledLearnable Behavioral Control (LBC) to address the limitation, which a) enablesa significantly enlarged behavior selection space via formulating a hybridbehavior mapping from all policies; b) constructs a unified learnable processfor behavior selection. We introduce LBC into distributed off-policyactor-critic methods and achieve behavior control via optimizing the selectionof the behavior mappings with bandit-based meta-controllers. Our agents haveachieved 10077.52% mean human normalized score and surpassed 24 human worldrecords within 1B training frames in the Arcade Learning Environment, whichdemonstrates our significant state-of-the-art (SOTA) performance withoutdegrading the sample efficiency.

Learnable Behavior Control: Breaking Atari Human World Records via Sample-Efficient Behavior Selection