Yutaro Yamada

I am a research scientist at Google DeepMind in London. Previously I was a research scientist at Sakana AI. Before that, I completed my PhD in Statistics and Data Science at Yale. My current interests include LLM agents, open-endedness, self-improving AI, and automated scientific discovery.

Research

• The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search
  Yutaro Yamada*, Robert Tjarko Lange*, Cong Lu*, Shengran Hu, Chris Lu, Jakob Foerster, Jeff Clune, David Ha
  Preprint, 2025 [Code] [Blog Post] [ICLR2025 Workshop Experiment] [Slides for NLP Colloquium (in Japanese)]
  Invited talks at ICLR 2025 ICBINB Workshop, PyTorch Conference 2025, MIT FutureTech Conference 2025, ICONIP 2025, and NeurIPS 2025 Social.

• L3GO: Language Agents with Chain-of-3D-Thoughts for Generating Unconventional Objects
  Yutaro Yamada, Khyathi Chandu, Bill Yuchen Lin, Jack Hessel, Ilker Yildirim, Yejin Choi
  NAACL 2025, System Demonstration track [Demo Video]

• Three-Dimensional Reconstruction Pre-Training as a Prior to Improve Robustness to Adversarial Attacks and Spurious Correlation
  Yutaro Yamada, Fred Weiying Zhang, Yuval Kluger, Ilker Yildirim
  Entropy 2024

• Evaluating Spatial Understanding of Large Language Models
  Yutaro Yamada*, Yihan Bao*, Andrew K. Lampinen, Jungo Kasai, Ilker Yildirim
  TMLR 2024

• Evaluating gpt-4 and chatgpt on japanese medical licensing examinations
  Jungo Kasai, Yuhei Kasai, Keisuke Sakaguchi, Yutaro Yamada, Dragomir Radev
  Preprint, 2023

• When are Lemons Purple? The Concept Association Bias of Vision-Language Models
  Yutaro Yamada*, Yingtian Tang*, Yoyo Zhang, Ilker Yildirim
  EMNLP 2023 [Slides]

• Support Recovery with Stochastic Gates: Theory and Application for Linear Models
  Soham Jana, Henry Li, Yutaro Yamada, Ofir Lindenbaum
  Signal Processing 2023

• Can Wikipedia Help Offline Reinforcement Learning?
  Machel Reid, Yutaro Yamada, Shixiang Shane Gu
  Preprint, 2022

• Does Robustness on ImageNet Transfer to Downstream Tasks?
  Yutaro Yamada, Mayu Otani
  CVPR 2022

• Exploiting 3D Shape Bias towards Robust Vision
  Yutaro Yamada, Sahand Negahban, Yuval Kluger, Ilker Yildirim
  NeurIPS 2021 Workshop SVRHM

• Feature Selection using Stochastic Gates
  Yutaro Yamada*, Ofir Lindenbaum*, Sahand Negahban, Yuval Kluger
  ICML 2020

• Understanding Adversarial Training: Increasing local stability of supervised models through robust optimization
  Uri Shaham, Yutaro Yamada, Sahand Negahban
  Neurocomputing 2018