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

Prithviraj Ammanabrolu, Ethan Tien, Matthew Hausknecht, Mark Riedl

Keywords: text-based games, reinforcement learning, exploration, intrinsic motivation, knowledge graphs, question answering, natural language processing

Abstract: Text-based games are long puzzles or quests, characterized by a sequence of sparse and potentially deceptive rewards. They provide an ideal platform to develop agents that perceive and act upon the world using a combinatorially sized natural language state-action space. Standard Reinforcement Learning agents are poorly equipped to effectively explore such spaces and often struggle to overcome bottlenecks---states that agents are unable to pass through simply because they do not see the right action sequence enough times to be sufficiently reinforced. We introduce Q*BERT, an agent that learns to build a knowledge graph of the world by answering questions, which leads to greater sample efficiency. To overcome bottlenecks, we further introduce MC!Q*BERT an agent that uses an knowledge-graph-based intrinsic motivation to detect bottlenecks and a novel exploration strategy to efficiently learn a chain of policy modules to overcome them. We present an ablation study and results demonstrating how our method outperforms the current state-of-the-art on nine text games, including the popular game, Zork, where, for the first time, a learning agent gets past the bottleneck where the player is eaten by a Grue.

One-sentence Summary: We present Q*BERT, an agent that systematically explores via an intrinsic motivation to learn a knowledge graph of the world.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2006.07409/code)

Reviewed Version (pdf): https://openreview.net/references/pdf?id=rA8gtHSFXC