# Grounding Language for Transfer in Deep Reinforcement Learning

arXiv: https://arxiv.org/abs/1708.00133 · HTML: https://mlsanity.com/paper/1708.00133/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Aug 2017.

## Built on by (most cited first, 5 of 5)

- [Cognitive Architectures for Language Agents](https://mlsanity.com/paper/2309.02427/) (2023)
- [Boosting Search Engines with Interactive Agents](https://mlsanity.com/paper/2109.00527/) (2021)
- [Teachable Reinforcement Learning via Advice Distillation](https://mlsanity.com/paper/2203.11197/) (2022)
- [Ask Before You Act: Generalising to Novel Environments by Asking Questions](https://mlsanity.com/paper/2209.04665/) (2022)
- [Feudal Reinforcement Learning by Reading Manuals](https://mlsanity.com/paper/2110.06477/) (2021)

## Abstract (arXiv)

In this paper, we explore the utilization of natural language to drive transfer for reinforcement learning (RL). Despite the wide-spread application of deep RL techniques, learning generalized policy representations that work across domains remains a challenging problem. We demonstrate that textual descriptions of environments provide a compact intermediate channel to facilitate effective policy transfer. Specifically, by learning to ground the meaning of text to the dynamics of the environment such as transitions and rewards, an autonomous agent can effectively bootstrap policy learning on a new domain given its description. We employ a model-based RL approach consisting of a differentiable planning module, a model-free component and a factorized state representation to effectively use entity descriptions. Our model outperforms prior work on both transfer and multi-task scenarios in a variety of different environments. For instance, we achieve up to 14% and 11.5% absolute improvement over previously existing models in terms of average and initial rewards, respectively.