Deep Reinforcement Learning with a Natural Language Action Space

2015/11/14 by Ji He, He, Ji, Jianshu Chen +11 · 7 citations
Computer Science · #Topic Modeling #Artificial Intelligence in Games #Natural Language Processing Techniques

paper · pdf · doi:10.48550/arxiv.1511.04636

Abstract

This paper introduces a novel architecture for reinforcement learning with deep neural networks designed to handle state and action spaces characterized by natural language, as found in text-based games. Termed a deep reinforcement relevance network (DRRN), the architecture represents action and state spaces with separate embedding vectors, which are combined with an interaction function to approximate the Q-function in reinforcement learning. We evaluate the DRRN on two popular text games, showing superior performance over other deep Q-learning architectures. Experiments with paraphrased action descriptions show that the model is extracting meaning rather than simply memorizing strings of text.

Citations

• Language Understanding for Text-based Games Using Deep Reinforcement\n Learning

Cited by

• Document-editing Assistants and Model-based Reinforcement Learning as a Path to Conversational AI
• Sample-Efficient Model-based Actor-Critic for an Interactive Dialogue\n Task
• Dual-Scale World Models for LLM Agents Towards Hard-Exploration Problems
• End-to-End Goal-Driven Web Navigation
• Parameterized Indexed Value Function for Efficient Exploration in Reinforcement Learning
• Reinforcement Learning with Function-Valued Action Spaces for Partial Differential Equation Control
• Counting to Explore and Generalize in Text-based Games

Related

• Playing Text-Adventure Games with Graph-Based Deep Reinforcement Learning
• Deep Reinforcement Learning: An Overview
• A Survey of Reinforcement Learning Informed by Natural Language
• Transformer Based Reinforcement Learning For Games
• Language Understanding for Text-based Games Using Deep Reinforcement\n Learning