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

Very large state spaces with a sparse reward signal are difficult to explore.The lack of a sophisticated guidance results in a poor performance for numerousreinforcement learning algorithms. In these cases, the commonly used randomexploration is often not helpful. The literature shows that this kind ofenvironments require enormous efforts to systematically explore large chunks ofthe state space. Learned state representations can help here to improve thesearch by providing semantic context and build a structure on top of the rawobservations. In this work we introduce a novel time-myopic staterepresentation that clusters temporal close states together while providing atime prediction capability between them. By adapting this model to theGo-Explore paradigm (Ecoffet et al., 2021b), we demonstrate the first learnedstate representation that reliably estimates novelty instead of using thehand-crafted representation heuristic. Our method shows an improved solutionfor the detachment problem which still remains an issue at the Go-ExploreExploration Phase. We provide evidence that our proposed method covers theentire state space with respect to all possible time trajectories withoutcausing disadvantageous conflict-overlaps in the cell archive. Analogous tonative Go-Explore, our approach is evaluated on the hard explorationenvironments MontezumaRevenge, Gravitar and Frostbite (Atari) in order tovalidate its capabilities on difficult tasks. Our experiments show thattime-myopic Go-Explore is an effective alternative for the domain-engineeredheuristic while also being more general. The source code of the method isavailable on GitHub.

Time-Myopic Go-Explore: Learning A State Representation for the Go-Explore Paradigm