Hao Zhu
Carnegie Mellon University
March 27, 2024
10:00am
3780 WEB
Finding Common Ground: Cooperation, Communication, Embodiment
Abstract
Intelligent agents are increasingly playing important social roles in human society, fueled by advancements in language, vision, and decision-making technologies. However, we are curious about whether machines are socially intelligent enough to interact with humans in the real world. In my talk, I will first explore this question through the lens of our latest research conducted in a simulated social environment, SOTOPIA. Our study sheds light on the complexities of developing cooperative AI, systems designed to work alongside humans in real-world settings through language. I will dive into our recent efforts to create a shared foundation for cooperative AI from three angles: (1) multi-agent interactive learning to boost the social intelligence and alignment of large language model agents; (2) training models to understand the mental states (Theory of Mind or ToM) of conversational partners, facilitating pragmatic communication; (3) building embodied and virtual environments, EXCALIBUR and WebArena, to investigate how to enable language agents to interact with the real world. The talk will conclude with a discussion of future pathways and challenges for this field, including grounding social intelligence in the physical world through human-robot interaction, safety and alignment of social agents, and an application field: cooperative AI for games.
Bio
Hao Zhu is a final-year PhD candidate in the Language Technologies Institute at Carnegie Mellon University. His research primarily focuses on real-world cooperative artificial intelligence (AI), which lies at the intersection of natural language processing (NLP), computer vision (CV), and cognitive science (CogSci). Specifically, he works on uncovering the role of language in real-world cooperation between humans and AI systems. He has published and presented his work at ACL, EMNLP, NAACL, ICML, ICLR, CVPR, and CogSci, including oral and spotlight presentations for six long papers at these conferences. He is also the lead co-organizer of the Workshop on Theory of Mind in Communicating Agents (ToM) at ICML 2023.