Autonomous Behavior
Autonomous behavior research focuses on enabling robots and computer agents to perform tasks independently, adapting to unpredictable environments and human interaction. Current efforts concentrate on developing hierarchical control architectures inspired by biological systems, leveraging reinforcement learning and large language models to achieve complex loco-manipulation and decision-making capabilities, often incorporating shared autonomy models to balance human oversight with robotic independence. This field is crucial for advancing robotics, human-computer interaction, and AI safety, with applications ranging from assistive technologies and industrial automation to autonomous vehicles and disaster response.
Papers
From Exploration to Mastery: Enabling LLMs to Master Tools via Self-Driven Interactions
Changle Qu, Sunhao Dai, Xiaochi Wei, Hengyi Cai, Shuaiqiang Wang, Dawei Yin, Jun Xu, Ji-Rong Wen
Agent S: An Open Agentic Framework that Uses Computers Like a Human
Saaket Agashe, Jiuzhou Han, Shuyu Gan, Jiachen Yang, Ang Li, Xin Eric Wang