Jiaxian Guo
Scholar

Jiaxian Guo

Google Scholar ID: wQgPocEAAAAJ
Google Research
Efficient Foundation ModelReinforcement LearningCausality
Citations & Impact
All-time
Citations
2,123
 
H-index
15
 
i10-index
15
 
Publications
20
 
Co-authors
15
list available
Resume (English only)
Academic Achievements
  • - Conference Reviewer: ICML 2021-2023, NeurIPS 2021-2022, ICLR 2021-2023, CVPR 2021-2023, ICCV 2021-2023, AAAI 2022
  • - Journal Reviewer: IEEE Journal of Cybernatics, IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), NeuroComputing, Machine Learning Journal (Springer), Transactions on Machine Learning Research
Research Experience
  • - Senior Research Scientist at Google Research Australia, main contributor to Gemma3n (open-sourced LLM for on-device scenarios) and Gemini-2.5 series and model design of the next generation Gemini
  • - Visiting Researcher at Matsuo Lab, The University of Tokyo, mentoring students exploring LLM-based agents, causality in LLMs, and multi-modal large language models
Education
  • - Ph.D.: Department of Computer Science, The University of Sydney, Advisors: Prof. Dacheng Tao and Dr. Mingming Gong
  • - PostDoc: The University of Tokyo, Advisors: Dr. Shane Gu and Prof. Yutaka Matsuo
  • - Bachelor’s Degree: Shanghai Jiao Tong University, Leader of Text Generation Group in Apex Lab, Undergraduate Research Advisors: Prof. Yong Yu, Prof. Weinan Zhang, and Prof. Jun Wang
Background
  • - Research Interests: Machine learning, particularly in causal learning, reinforcement learning, and domain adaptation methods, as well as their applications in sequential decision making, computer vision, and natural language processing
  • - Professional Field: Enhancing the inference efficiency of large foundation models and proposing efficient model architectures
  • - Brief Introduction: Currently a Senior Research Scientist at Google Research Australia, dedicated to enhancing the inference efficiency of large foundation models and proposing efficient model architectures.
Miscellany
  • Looking for outstanding and highly motivated students or collaborators to work together in the above directions