DoReMi: Grounding Language Model by Detecting and Recovering from Plan-Execution Misalignment

no code implementations • 1 Jul 2023 • Yanjiang Guo, Yen-Jen Wang, Lihan Zha, Zheyuan Jiang, Jianyu Chen

Large language models (LLMs) encode a vast amount of semantic knowledge and possess remarkable understanding and reasoning capabilities.

Language Modelling Question Answering +1

Decentralized Motor Skill Learning for Complex Robotic Systems

no code implementations • 30 Jun 2023 • Yanjiang Guo, Zheyuan Jiang, Yen-Jen Wang, Jingyue Gao, Jianyu Chen

Experiments on quadruped and humanoid robots demonstrate that the learned policy is robust against local motor malfunctions and can be transferred to new tasks.

Reinforcement Learning (RL)

Reinforcement learning with Demonstrations from Mismatched Task under Sparse Reward

no code implementations • 3 Dec 2022 • Yanjiang Guo, Jingyue Gao, Zheng Wu, Chengming Shi, Jianyu Chen

In this paper, we consider the case where the target task is mismatched from but similar with that of the expert.

reinforcement-learning Reinforcement Learning (RL) +1

Zero-Shot Policy Transfer with Disentangled Task Representation of Meta-Reinforcement Learning

no code implementations • 1 Oct 2022 • Zheng Wu, Yichen Xie, Wenzhao Lian, Changhao Wang, Yanjiang Guo, Jianyu Chen, Stefan Schaal, Masayoshi Tomizuka

Experimental results demonstrate that our proposed method achieves policy generalization to unseen compositional tasks in a zero-shot manner.

Disentanglement Meta Reinforcement Learning +2