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 (RL) agents are often unable to generalise well toenvironment variations in the state space that were not observed duringtraining. This issue is especially problematic for image-based RL, where achange in just one variable, such as the background colour, can change manypixels in the image. The changed pixels can lead to drastic changes in theagent's latent representation of the image, causing the learned policy to fail.To learn more robust representations, we introduce TEmporal Disentanglement(TED), a self-supervised auxiliary task that leads to disentangled imagerepresentations exploiting the sequential nature of RL observations. We findempirically that RL algorithms utilising TED as an auxiliary task adapt morequickly to changes in environment variables with continued training compared tostate-of-the-art representation learning methods. Since TED enforces adisentangled structure of the representation, our experiments also show thatpolicies trained with TED generalise better to unseen values of variablesirrelevant to the task (e.g. background colour) as well as unseen values ofvariables that affect the optimal policy (e.g. goal positions).

Temporal Disentanglement of Representations for Improved Generalisation in Reinforcement Learning