RTFM: Generalising to Novel Environment Dynamics via Reading

By Victor Zhong, Tim Rocktaschel, Edward Grefenstette. 2019

55 citations recorded by the source corpus.

Obtaining policies that can generalise to new environments in reinforcement learning is challenging. In this work, we demonstrate that language understanding via a reading policy learner is a promising vehicle for generalisation to new environments. We propose a grounded policy learning problem, Read to Fight Monsters (RTFM), in which the agent must jointly reason over a language goal, relevant dynamics described in a document, and environment observations. We procedurally generate environment dynamics and corresponding language descriptions of the dynamics, such that agents must read to understand new environment dynamics instead of memorising any particular information. In addition, we propose txt2$\pi$, a model that captures three-way interactions between the goal, document, and observations. On RTFM, txt2$\pi$ generalises to new environments with dynamics not seen during training via reading. Furthermore, our model outperforms baselines such as FiLM and language-conditioned CNNs on RTFM. Through curriculum learning, txt2$\pi$ produces policies that excel on complex RTFM tasks requiring several reasoning and coreference steps.

Subfields: Artificial Intelligence, Natural Language Processing, Reinforcement Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

In Reinforcement Learning

Citations in the corpus

Human Instruction-Following with Deep Reinforcement Learning via Transfer-Learning from Text (2020, methodology): “To make the approach feasible, many studies thus employ a synthetic language that is generated on demand from templates by the environment simulator [Chevalier-Boisvert et al., 2018, Jiang et al., 2019, Yu et al., 2018b,a, Zhong et al., 2020].”
Language Conditioned Imitation Learning Over Unstructured Data (2020, background): “Typical studies involve 2D observation spaces (e.g. games [24, 32, 27, 12] and gridworlds [52]), simpliﬁed actuators, (e.g. binary pick and place primitives [18, 48]), or synthetic predeﬁned language [17, 6, 20, 54].”
Do Embodied Agents Dream of Pixelated Sheep?: Embodied Decision Making using Language Guided World Modelling (2023, methodology): “Textual knowledge can be used to improve generalization in reinforcement learning through environment descriptions (Branavan et al., 2011; Zhong et al., 2020; Hanjie et al., 2021) or language instructions (Chevalier-Boisvert et al., 2019; Anderson et al., 2018; Ku et al., 2020; Shridhar et al.,…”