Timezone: »

 
Poster
Improving Policy Learning via Language Dynamics Distillation
Victor Zhong · Jesse Mu · Luke Zettlemoyer · Edward Grefenstette · Tim Rocktäschel

Wed Nov 30 09:00 AM -- 11:00 AM (PST) @ Hall J #912

Recent work has shown that augmenting environments with language descriptions improves policy learning. However, for environments with complex language abstractions, learning how to ground language to observations is difficult due to sparse, delayed rewards. We propose Language Dynamics Distillation (LDD), which pretrains a model to predict environment dynamics given demonstrations with language descriptions, and then fine-tunes these language-aware pretrained representations via reinforcement learning (RL). In this way, the model is trained to both maximize expected reward and retain knowledge about how language relates to environment dynamics. On SILG, a benchmark of five tasks with language descriptions that evaluate distinct generalization challenges on unseen environments (NetHack, ALFWorld, RTFM, Messenger, and Touchdown), LDD outperforms tabula-rasa RL, VAE pretraining, and methods that learn from unlabeled demonstrations in inverse RL and reward shaping with pretrained experts. In our analyses, we show that language descriptions in demonstrations improve sample-efficiency and generalization across environments, and that dynamics modeling with expert demonstrations is more effective than with non-experts.

Author Information

Victor Zhong (University of Washington)
Jesse Mu (Stanford University)
Luke Zettlemoyer (University of Washington and Facebook)
Edward Grefenstette (Cohere & University College London)
Tim Rocktäschel (University College London, Facebook AI Research)

Tim is a Researcher at Facebook AI Research (FAIR) London, an Associate Professor at the Centre for Artificial Intelligence in the Department of Computer Science at University College London (UCL), and a Scholar of the European Laboratory for Learning and Intelligent Systems (ELLIS). Prior to that, he was a Postdoctoral Researcher in Reinforcement Learning at the University of Oxford, a Junior Research Fellow in Computer Science at Jesus College, and a Stipendiary Lecturer in Computer Science at Hertford College. Tim obtained his Ph.D. from UCL under the supervision of Sebastian Riedel, and he was awarded a Microsoft Research Ph.D. Scholarship in 2013 and a Google Ph.D. Fellowship in 2017. His work focuses on reinforcement learning in open-ended environments that require intrinsically motivated agents capable of transferring commonsense, world and domain knowledge in order to systematically generalize to novel situations.

More from the Same Authors