Boqin Yuan

b4yuan[at]ucsd.edu

I’m a Master student in Computer Science at UC San Diego, exploring the intersection of Machine Learning and Software Engineering. At UCSD, I work with STABLE Lab (Prof. Jishen Zhao) on AI Agent and ML systems research.

I was a full-time Machine Learning Engineer at CambioML (YC S23) from 2024-2025, working on vision language models (AnyParser) and Computer Use Agent (CUA) systems (Energent.ai).

Previously studied Mathematics & Computer Science and Statistics at University of Illinois at Urbana-Champaign (UIUC). I bridge research ideas and production-ready engineering.

Research Interests: Machine Learning, ML Systems, AI Agent, Reinforcement Learning, AI Safety

Email GitHub LinkedIn Google Scholar CV

News

Apr 27, 2026	ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation released on arXiv. Check out the website and code!
Mar 03, 2026	AMA-Bench accepted to ICLR 2026 Workshop MemAgents.
Mar 03, 2026	Diagnosing Retrieval vs. Utilization Bottlenecks accepted to ICLR 2026 MemAgents.
Feb 27, 2026	AMA-Bench for agent trajectory memory released on arXiv. Check out our benchmark!

Selected Publications

ICLR-W

AMA-Bench: Evaluating Long-Horizon Memory for Agentic Applications

Yujie Zhao, Boqin Yuan, Junbo Huang, Haocheng Yuan, Zhongming Yu, Haozhou Xu, and 6 more authors

In ICLR 2026 Workshop MemAgents (Oral), 2026

Paper Bib Website

@inproceedings{zhao2026amabench,
  title = {AMA-Bench: Evaluating Long-Horizon Memory for Agentic Applications},
  author = {Zhao, Yujie and Yuan, Boqin and Huang, Junbo and Yuan, Haocheng and Yu, Zhongming and Xu, Haozhou and Hu, Lanxiang and Shankarampeta, Abhilash and Huang, Zimeng and Ni, Wentao and Tian, Yuandong and Zhao, Jishen},
  booktitle = {ICLR 2026 Workshop MemAgents (Oral)},
  year = {2026},
}

arXiv

ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation

Boqin Yuan, Renchu Song, Yue Su, Sen Yang, and Jing Qin

arXiv preprint arXiv:2604.23853, 2026

Paper Bib Code Website

@article{yuan2026clawtrace,
  title = {ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation},
  author = {Yuan, Boqin and Song, Renchu and Su, Yue and Yang, Sen and Qin, Jing},
  journal = {arXiv preprint arXiv:2604.23853},
  year = {2026},
}

DAC

PRO-V-R1: Reasoning Enhanced Programming Agent for RTL Verification

Yujie Zhao, Zhijing Wu, Boqin Yuan, Zhongming Yu, Hejia Zhang, Wentao Ni, and 3 more authors

In Design Automation Conference (DAC), 2026

Paper Bib

@inproceedings{zhao2026provr1,
  title = {PRO-V-R1: Reasoning Enhanced Programming Agent for RTL Verification},
  author = {Zhao, Yujie and Wu, Zhijing and Yuan, Boqin and Yu, Zhongming and Zhang, Hejia and Ni, Wentao and Ho, Chia-Tung and Ren, Haoxing and Zhao, Jishen},
  booktitle = {Design Automation Conference (DAC)},
  year = {2026},
}

ICLR-W

Diagnosing Retrieval vs. Utilization Bottlenecks in LLM Agent Memory

Boqin Yuan, Yue Su, and Kun Yao

In ICLR 2026 Workshop MemAgents (Poster), 2026

Paper Bib

@inproceedings{yuan2026diagnosing,
  title = {Diagnosing Retrieval vs. Utilization Bottlenecks in LLM Agent Memory},
  author = {Yuan, Boqin and Su, Yue and Yao, Kun},
  booktitle = {ICLR 2026 Workshop MemAgents (Poster)},
  year = {2026},
}