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) in long horizon and sparse reward tasks isnotoriously difficult and requires a lot of training steps. A standard solutionto speed up the process is to leverage additional reward signals, shaping it tobetter guide the learning process. In the context of language-conditioned RL,the abstraction and generalisation properties of the language input provideopportunities for more efficient ways of shaping the reward. In this paper, weleverage this idea and propose an automated reward shaping method where theagent extracts auxiliary objectives from the general language goal. Theseauxiliary objectives use a question generation (QG) and question answering (QA)system: they consist of questions leading the agent to try to reconstructpartial information about the global goal using its own trajectory. When itsucceeds, it receives an intrinsic reward proportional to its confidence in itsanswer. This incentivizes the agent to generate trajectories whichunambiguously explain various aspects of the general language goal. Ourexperimental study shows that this approach, which does not require engineerintervention to design the auxiliary objectives, improves sample efficiency byeffectively directing exploration.

EAGER: Asking and Answering Questions for Automatic Reward Shaping in Language-guided RL