ICML 2026
Good SFT Optimizes for SFT, Better SFT Prepares for Reinforcement Learning
Dylan Zhang, Yufeng Xu, Haojin Wang, Qingzhi Chen, Hao Peng
About Me
I am Dylan, a Ph.D. Student in Computer Science at the University of Illinois Urbana-Champaign (UIUC) 🌽, advised by Prof. Hao Peng. I am a proud member of our lab, officially named ALTA.
I’m broadly interested in language model post-training. My previous work has focused on the data-centric aspects of post-training. Moving forward, I’m eager to explore new directions such as:
- 🚀 Offline-to-online reinforcement learning
- 💡 Incentivizing proactive reasoning behaviors for knowledge discovery
- 🤖 Rapid deployment of foundation model–based agents in novel environments
I’m also deeply excited about the growing role of foundation models in scientific discovery—and the rich post-training algorithmic challenges that come with it.
Looking for Opportunities
I am looking for Research Internship opportunities for Summer 2026.
Education
🎓
Ph.D. in Computer Science
University of Illinois Urbana-Champaign 🌽 🐿️
2022 - Present
Advisor: Prof. Hao Peng
News
Feb 2026
ICML 2026
Our paper Good SFT Optimizes for SFT, Better SFT Prepares for Reinforcement Learning was accepted to ICML 2026! 🎉
Sep 2025
GRAPE
was accepted to NeurIPS 2025 Spotlight! 🎉
May 2025
Started my Student Researcher journey with Google. 🚀
Ongoing Works
PEAR
View project page
Internship Experience
Student Researcher
Mountain View, CA
Research Intern
Microsoft Research
Redmond, WA
Research Intern
Microsoft Research
Redmond, WA
Selected Works
26
2026
25
2025
The Best Instruction-Tuning Data are Those That Fit
Dylan Zhang, Qirun Dai, Hao Peng
Improving Influence-based Instruction Tuning Data Selection for Balanced Learning of Diverse Capabilities
Qirun Dai, Dylan Zhang, Jiaqi W. Ma, Hao Peng
Entropy-Regularized Process Reward Model
Hanning Zhang*, Pengcheng Wang*, Shizhe Diao*, Yong Lin, Rui Pan, Dylan Zhang, Pavlo Molchanov, Tong Zhang
ScaleBiO: Scalable Bilevel Optimization for LLM Data Reweighting
Dylan Zhang*, Rui Pan*, Hanning Zhang*, Xingyuan Pan*, Minrui Xu, Jipeng Zhang, Renjie Pi, Xiaoyu Wang, Tong Zhang (*equal contrib)
24
2024
Only-IF: Revealing the Decisive Effect of Instruction Diversity on Generalization
Dylan Zhang, Justin Wang, Francois Charton
SciCode: A Research Coding Benchmark Curated by Scientists
Minyang Tian*, Luyu Gao*, Dylan Zhang, Xinan Chen, … (multi-institution collaboration)
23
2023
Making Large Language Models Better Reasoners with Step-Aware Verifier
Yifei Li, Zeqi Lin, Dylan Zhang, Qiang Fu, Bei Chen, Jian-Guang Lou, Weizhu Chen
21
2021
Pre-training Co-evolutionary Protein Representation via a Pairwise Masked Language Model
Liang He, Dylan Zhang, Lijun Wu, Huanhuan Xia, Fusong Ju, He Zhang, Siyuan Liu, …, Tie-Yan Liu