Yutaro Yamada is a Research Scientist at Sakana AI. He received his PhD in Statistics and Data Science from Yale University and worked at Meta and the Allen Institute for AI as a research intern. At Sakana AI, he co-led the development of AI Scientist-v2. His research interests include LLM agents, open-endedness, self-improving AI, and automated scientific discovery.
[Webサイト]
概要：This talk surveys the emerging landscape of AI systems for scientific research, from coding agents and AI co-scientists to closed-loop discovery systems. I will use The AI Scientist, our recent Nature publication, as a case study for understanding how far current LLM-based systems can be pushed toward autonomous research when combined with tools, execution environments, search, and review mechanisms. I will discuss challenges in problem design, verification, hallucination, and transparency around these systems. The talk will conclude with open questions about how future AI Scientist systems should be evaluated, governed, and integrated into scientific practice.
[論文] (Nature)
※トークは日本語です。
参加用のZoom URLはメーリングリストで配信しています。
ほかの宣伝投稿はありません。いつでも抜けられます。