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

Learning a diverse set of skills by interacting with an environment withoutany external supervision is an important challenge. In particular, obtaining agoal-conditioned agent that can reach any given state is useful in manyapplications. We propose a novel method for training such a goal-conditionedagent without any external rewards or any domain knowledge. We use random walkto train a reachability network that predicts the similarity between twostates. This reachability network is then used in building goal memorycontaining past observations that are diverse and well-balanced. Finally, wetrain a goal-conditioned policy network with goals sampled from the goal memoryand reward it by the reachability network and the goal memory. All thecomponents are kept updated throughout training as the agent discovers andlearns new goals. We apply our method to a continuous control navigation androbotic manipulation tasks.

Walk the Random Walk: Learning to Discover and Reach Goals Without Supervision