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

Reinforcement learning algorithms can acquire policies for complex tasksautonomously. However, the number of samples required to learn a diverse set ofskills can be prohibitively large. While meta-reinforcement learning methodshave enabled agents to leverage prior experience to adapt quickly to new tasks,their performance depends crucially on how close the new task is to thepreviously experienced tasks. Current approaches are either not able toextrapolate well, or can do so at the expense of requiring extremely largeamounts of data for on-policy meta-training. In this work, we present modelidentification and experience relabeling (MIER), a meta-reinforcement learningalgorithm that is both efficient and extrapolates well when faced without-of-distribution tasks at test time. Our method is based on a simpleinsight: we recognize that dynamics models can be adapted efficiently andconsistently with off-policy data, more easily than policies and valuefunctions. These dynamics models can then be used to continue training policiesand value functions for out-of-distribution tasks without usingmeta-reinforcement learning at all, by generating synthetic experience for thenew task.

Meta-Reinforcement Learning Robust to Distributional Shift via Model Identification and Experience Relabeling