An Overview of Natural Language State Representation for Reinforcement Learning
arXiv:2007.09774
Abstract
A suitable state representation is a fundamental part of the learning process in Reinforcement Learning. In various tasks, the state can either be described by natural language or be natural language itself. This survey outlines the strategies used in the literature to build natural language state representations. We appeal for more linguistically interpretable and grounded representations, careful justification of design decisions and evaluation of the effectiveness of different approaches.
arXiv commentAccepted to the ICML 2020 Workshop on Language in Reinforcement Learning (LaReL). 4 pages
References in corpus (22)
- A Deep Reinforced Model for Abstractive Summarization
- Learning to Communicate with Deep Multi-Agent Reinforcement Learning
- Sequence Level Training with Recurrent Neural Networks
- Deep Reinforcement Learning for Dialogue Generation
- State Representation Learning for Control: An Overview
- Finding Approximate POMDP solutions Through Belief Compression
- Grounded Language Learning in a Simulated 3D World
- End-to-end LSTM-based dialog control optimized with supervised and reinforcement learning
- Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning
- Continuously Learning Neural Dialogue Management
- Gated-Attention Architectures for Task-Oriented Language Grounding
- BBQ-Networks: Efficient Exploration in Deep Reinforcement Learning for Task-Oriented Dialogue Systems
- From Language to Goals: Inverse Reinforcement Learning for Vision-Based Instruction Following
- End-to-End Task-Completion Neural Dialogue Systems
- Counting to Explore and Generalize in Text-based Games
- From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood
- Beating Atari with Natural Language Guided Reinforcement Learning
- Selecting the State-Representation in Reinforcement Learning
- The Thing That We Tried Didn't Work Very Well : Deictic Representation in Reinforcement Learning
- Enhancing Text-based Reinforcement Learning Agents with Commonsense Knowledge
- Language Expansion In Text-Based Games
- Tackling Error Propagation through Reinforcement Learning: A Case of Greedy Dependency Parsing