About Me

Hi! I’m Sebastian Zhao, a ml researcher. I’m currently doing research in pallas working with Coleman Hooper under Professor Kurt Keutzer. I like cooking, creating plated desserts, thrifting, playing guitar, and skateboarding.

Research Interest

I’m broadly interested in efficient ML. My recent work has been on model compression and long context inference. I’m particularly interested in optimizing across models, systems, and hardware with the goal of democratizing ML! I’ve previously worked on coding assistants, robotics, and AI ethics.

Publications

• CodingGenie: A Proactive LLM-Powered Programming Assistant FSE Demo 2025
  Sebastian Zhao, Alan Zhu, Hussein Mozannar, David Sontag, Ameet Talwalkar, Valerie Chen

• The Ingredients for Robotic Diffusion Transformers ICRA 2025
  Sudeep Dasari, Oier Mees, Sebastian Zhao, Mohan Kumar Srirama, Sergey Levine

• Need Help? Designing Proactive AI Assistants for Programming CHI 2025
  Valerie Chen, Alan Zhu, Sebastian Zhao, Hussein Mozannar, David Sontag, Ameet Talwalkar

• Engagement, User Satisfaction, and the Amplification of Divisive Content on Social Media PNAS Nexus
  Smitha Milli, Micah Carroll, Yike Wang, Sashrika Pandey, Sebastian Zhao, Anca D. Dragan

• The RealHumanEval: Evaluating Large Language Models’ Abilities to Support Programmers TMLR 2025
  Hussein Mozannar, Valerie Chen, Mohammed Alsobay, Subhro Das, Sebastian Zhao, Dennis Wei, Manish Nagireddy, Prasanna Sattigeri, Ameet Talwalkar, David Sontag

• One Less Reason for Filter Pruning: Gaining Free Adversarial Robustness with Structured Grouped Kernel Pruning NeurIPS 2023
  Shaochen Zhong, Zaichuan You, Jiamu Zhang, Sebastian Zhao*, Zachary LeClaire, Zirui Liu, Daochen Zha, Vipin Chaudhary, Shuai Xu, Xia Hu

News

• 9/19/25 – Multipole Attention for Efficient Long Context Reasoning was accepted into NeurIPS 2025
• 8/20/25 – I will be doing a research assistantship under Professor Kurt Keutzer for the next year.
• 3/10/25 – CodingGenie: A Proactive LLM-Powered Programming Assistant was accepted into FSE Demo 2025
• 1/28/25 – Engagement, User Satisfaction, and the Amplification of Divisive Content on Social Media was accepted into PNAS Nexus
• 1/27/25 – The Ingredients for Robotic Diffusion Transformers was accepted into ICRA 2025
• 1/16/25 – Need Help? Designing Proactive AI Assistants for Programming was accepted into CHI 2025
• 12/28/24 – The RealHumanEval: Evaluating Large Language Models’ Abilities to Support Programmers was accepted into TMLR 2025