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 are long puzzles or quests, characterized by a sequence ofsparse and potentially deceptive rewards. They provide an ideal platform todevelop agents that perceive and act upon the world using a combinatoriallysized natural language state-action space. Standard Reinforcement Learningagents are poorly equipped to effectively explore such spaces and oftenstruggle to overcome bottlenecks---states that agents are unable to passthrough simply because they do not see the right action sequence enough timesto be sufficiently reinforced. We introduce Q*BERT, an agent that learns tobuild a knowledge graph of the world by answering questions, which leads togreater sample efficiency. To overcome bottlenecks, we further introduceMC!Q*BERT an agent that uses an knowledge-graph-based intrinsic motivation todetect bottlenecks and a novel exploration strategy to efficiently learn achain of policy modules to overcome them. We present an ablation study andresults demonstrating how our method outperforms the current state-of-the-arton nine text games, including the popular game, Zork, where, for the firsttime, a learning agent gets past the bottleneck where the player is eaten by aGrue.

How to Avoid Being Eaten by a Grue: Structured Exploration Strategies for Textual Worlds