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

Text-based games (TBGs) have become a popular proving ground for thedemonstration of learning-based agents that make decisions in quasi real-worldsettings. The crux of the problem for a reinforcement learning agent in suchTBGs is identifying the objects in the world, and those objects' relations withthat world. While the recent use of text-based resources for increasing anagent's knowledge and improving its generalization have shown promise, we positin this paper that there is much yet to be learned from visual representationsof these same worlds. Specifically, we propose to retrieve images thatrepresent specific instances of text observations from the world and train ouragents on such images. This improves the agent's overall understanding of thegame 'scene' and objects' relationships to the world around them, and thevariety of visual representations on offer allow the agent to generate a bettergeneralization of a relationship. We show that incorporating such imagesimproves the performance of agents in various TBG settings.

Eye of the Beholder: Improved Relation Generalization for Text-based Reinforcement Learning Agents