Welcome to my homepage!

I am a Computer Science undergraduate at the University of Minnesota Twin Cities. I work on AI agents and large language models โ€” self-evolving agents, agent harnesses, and reinforcement learning for long-horizon terminal tasks. I build agent infrastructure and the benchmarks used to evaluate it.

My current research interests include:

(1) Self-evolving agents. Agents that improve from their own execution traces โ€” synthesizing harder tasks, verifying their own outcomes, and feeding what survives back in as training signal.

(2) Agent harnesses. The scaffolding around the model: tool interfaces, context management, and execution environments. How much of a model's capability actually shows up on a task often depends on the harness as much as the weights.

(3) Post-training. Turning a base model into a dependable agent โ€” supervised fine-tuning, preference optimization, and the data synthesis pipelines that feed them.

(4) Reinforcement learning. Reward design and credit assignment for agents acting over long horizons, where a single pass/fail at the end says very little about where the run actually went wrong.

Diagram of the recursive synthesis loop: synthesize a harder task, verify it, and train on solved rollouts
The recursive synthesis loop from our latest work: the agent synthesizes a harder task, iterates until it is oracle-valid, and keeps improving by training on the solved rollouts.

Outside research, I have competed in algorithm contests for about ten years, from NOI and USACO to the ICPC World Finals, and now coach the UMN team. Feel free to reach out if you want to talk about agents, evaluation, or contest problems!

๐Ÿ”ฅ News

  • 2026.08: ๐ŸŽ‰๐ŸŽ‰ Our paper Recursive Synthesis for Long-Horizon Terminal Tasks is now available on arXiv.
  • 2026.07: ๐ŸŽ‰๐ŸŽ‰ We released Long-Horizon-Terminal-Bench, a benchmark for evaluating agents on extended terminal tasks.
  • 2026.07: ๐Ÿ’ผ Completed my second software engineering internship at VecML in Bellevue, WA, working on LLM agent infrastructure.
  • 2026.03: ๐Ÿ† Qualified for the ICPC World Finals โ€” ranked 9th in North America at ICPC NAC, the first UMN team to qualify in 10 years.
  • 2025.11: ๐Ÿฅ‡ Won Gold at the ICPC North Central NA Regional Contest, our second consecutive regional gold.
  • 2025.08: ๐Ÿ’ผ Completed a software engineering internship at VecML in Bellevue, WA, building an AI data analysis platform.
  • 2024.11: ๐Ÿฅ‡ Won Gold at the ICPC North Central NA Regional Contest.
  • 2024.09: ๐ŸŽ“ Started my Bachelor's in Computer Science at the University of Minnesota Twin Cities.

๐Ÿ“ Publications

You can also find my papers on Google Scholar.

arXiv 2026 Recursive verified synthesis loop for generating increasingly difficult terminal tasks

Recursive Synthesis for Long-Horizon Terminal Tasks

Zhongzhi Li, Yucheng Shi, Zongxia Li, Ruhan Wang, Anhao Li, Zixun Huang, Junyao Yang, Lei Ke, Ninghao Liu, Haitao Mi, Leowei Liang

Paper Project

  • A recursive, verification-driven pipeline that synthesizes progressively harder terminal tasks, giving agents a training and evaluation signal that scales with their capability.
arXiv 2026 Long-Horizon Terminal-Bench leaderboard showing model performance

Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading

Zongxia Li, Zhongzhi Li, Yucheng Shi, Ruhan Wang, Junyao Yang, Zhichao Liu, Xiyang Wu, Anhao Li, Yue Yu, Ninghao Liu, Lichao Sun, Haotao Mi, Leowei Liang

Paper Project

  • A benchmark for agents on extended terminal tasks, replacing pass/fail checks with dense reward-based grading so partial progress is measured rather than discarded.

๐ŸŽ– Honors and Awards

  • 2026.03 ICPC World Finalist 2026, ranked 9th in North America at ICPC NAC; the first UMN team to qualify in 10 years.
  • 2025.11, 2024.11 Gold Medal, ICPC North Central NA Regional Contest, in two consecutive years.
  • 6ร— First Prize, NOIP / CSP.
  • 2021.12 USACO Platinum, ranked among the global top 30.
  • 2021.07 Bronze Medal, National Olympiad in Informatics (NOI).
  • 2021.01 Silver Medal, National Olympiad in Informatics Winter Camp.

๐Ÿ“– Education

๐Ÿ’ป Internships

  • 2026.05 โ€“ 2026.07, Software Engineer Intern, VecML Inc. (AI startup), Bellevue, WA. LLM agent infrastructure and evaluation.
  • 2025.06 โ€“ 2025.08, Software Engineer Intern, VecML Inc. (AI startup), Bellevue, WA. AI data analysis platform.

๐Ÿง‘โ€๐Ÿซ Teaching & Service

  • 2026 โ€“ 2027, ICPC Coach, University of Minnesota: practice contests, training plans, and problem-set design.
  • Ongoing, Competitive programming mentoring: contest editorials and training notes on my blog, and mentoring students moving from introductory algorithms to ICPC-level problems.

๐Ÿช„ Misc

I write contest write-ups, technical notes, and the occasional reflection on my blog. Every post is available in English and Chinese. Recent posts:

Email is the fastest way to reach me: li003978@umn.edu or andy.anhao.li@gmail.com.