Hey, I’m Bohan Yao. I’m currently an undergraduate at the University of Washington majoring in computer science and mathematical statistics. I’m also currently working at ServiceNow part-time, where I’m working on multimodal agentic reasoning systems. I previously interned at ServiceNow during 2024 and 2025, where I worked on LLM post-training and agentic systems.

I work on natural language processing (NLP) research, and my primary research interest is in developing systems that can tackle challenging reasoning tasks across different problem domains, such as math, coding, science, and involving diverse data modalities, such as text, images, charts. I have worked on research involving tool-augmented LLM frameworks and agentic reasoning systems.

I am currently searching for Fall 2026 PhD positions for LLM reasoning and agentic systems. Please feel free to reach out if you have any opportunities/information.

CV SoP

Publications

• ARM: Discovering Agentic Reasoning Modules for Generalizable Multi-Agent Systems
  Bohan Yao, Shiva Krishna Reddy Malay, Vikas Yadav
  The 5th Workshop on Mathematical Reasoning and AI @ NeurIPS (2025)
  PDF
• Diverse Multi-tool Aggregation with Large Language Models for Enhanced Math Reasoning
  Bohan Yao, Vikas Yadav
  EMNLP Findings (2025) & The 5th Workshop on Mathematical Reasoning and AI @ NeurIPS (2025)
  PDF

Contact

I am interested in collaborating on agentic systems and LLM reasoning research. If you have any interesting projects you think I would be a good fit for or want to talk for any other reasons, please reach out via email.