Ph.D. Student

Computer Science Department

University of Virginia

Email: {first_name}{last_name}@virginia.edu

Currently I’m working on:

• Post-training for LLM reasoning
• Synthetic data for LLM reasoning

X(Twitter) | Linkedin | CV

Github | Google Scholar

Hi!👋 I’m Xinyu (/ʃɪn.ju:/) Zhu, currently a Ph.D. student at the University of Virginia, advised by Prof. ****Yu Meng. I work on improving LLMs’ reasoning capability. I received my master degree at Tsinghua University, advised by Prof. ****Yujiu Yang.

✨News

• 🔥🔥🔥 [03/2026] Check out our new paper CHIMERA! We release a compact synthetic reasoning dataset of 9K expert-level problems across 8 disciplines, enabling a 4B model to match significantly larger reasoning models!
• [03/2026] Grateful to receive the Lambda's Research Grant ($5,000 cloud credits)!
• [09/2025] Two works on how negative-reward-only RL can improve LLM reasoning and transferring small RL-tuned models’ reasoning capability to large base model are accepted to NeurIPS 2025! ****🎉 See you in San Diego! 🤩
• [06/2025] Check out our new preprint and X thread on how negative-reward-only RL can improve LLM reasoning! 🚀
• [06/2025] Our AdaDecode is accepted at ICML 2025! 🎉
• [05/2025] I moved to Cupertino and started my internship at Apple AIML, looking forward to meeting new friends in the Bay Area! 🚀
• [03/2025] One survey on the honesty of LLMs is accepted at TMLR! 🎉
• [01/2025] One paper on evaluating LMM's reasoning via chart-to-code generation ****is accepted at ICLR 2025! 🎉
• Old news

👨‍🎓Education

• University of Virginia Ph.D. in Computer Science
• Tsinghua University Master in Electronic and Information Engineering
• Xidian University Bachelor in Electronic Science and Technology

📝Selected Works (* indicates equal contribution)

CHIMERA: Compact Synthetic Data for Generalizable LLM Reasoning

Preprint

Xinyu Zhu, Yihao Feng, Yanchao Sun, Xianzhi Du, Pingzhi Li, Olli Saarikivi, Yun Zhu, Yu Meng

[arxiv][dataset][models]