Mission
The Agentic AI Frontier Seminar advances research and practice on agentic AI, including safety, alignment, evaluation, tool augmentation, memory, long-horizon reasoning, and multi-agent collaboration.
About the Seminar
Format & Frequency
- Online. Talk 40–45 min + Q&A 10–20 min.
- Weekly by default (Pacific Time: 09:00–10:00 AM, PT). Special events may vary.
Audience
Researchers, engineers, and students in Agentic AI, RL, NLP, systems, security, and HCI.
Proposals & Collaboration
We welcome topic proposals / tutorials / postmortems / open-source releases. Please send a short abstract (≤ 300 words), speaker info, and preferred dates to shangding.gu@berkeley.edu, jinming@vt.edu.