Shuibai Zhang (张水柏) | University of Wisconsin-Madison

About Me

I am a second-year Computer Science Ph.D. student at the University of Wisconsin-Madison. My research focuses on the post-training stages of large foundation models, with the aim of making them more efficient and intelligent. Recently, I have become interested in diffusion language models, which offer inherent advantages such as parallel decoding and built-in error correction, and may hold the potential to reshape future paradigms of language modeling.

As an impact-driven researcher, I wish to build things that work and create real impact.

Projects

🔥 Open-dCoder: The First Fully Open Diffusion LLM for Code

A 0.5B diffusion coding model with pretraining pipeline, inference code, checkpoints, and evaluation suite.

📝 Notion Blog | 💻 GitHub

Open Source Project

Preprints

The Long-Horizon Task Mirage? Diagnosing Where and Why Agentic Systems Break

Xinyu Jessica Wang, Haoyue Bai, Yiyou Sun, Haorui Wang, Shuibai Zhang, Wenjie Hu, Mya Schroder, Bilge Mutlu, Dawn Song, Robert D Nowak

ArXiv

Expert-Choice Routing Enables Adaptive Computation in Diffusion Language Models

Shuibai Zhang*, Caspian Zhuang*, Chihan Cui*, Zhihan Yang, Fred Zhangzhi Peng, Yanxin Zhang, Haoyue Bai, Zack Jia, Yang Zhou, Guanhua Chen, Ming Liu

ArXiv equal contribution

Corrective Diffusion Language Model

Shuibai Zhang, Fred Zhangzhi Peng, Yiheng Zhang, Jin Pan, Grigorios G Chrysos

ArXiv

ReJump: A Tree-Jump Representation for Analyzing and Improving LLM Reasoning

Yuchen Zeng*, Shuibai Zhang*, Wonjun Kang*, et al., Dimitris Papailiopoulos, Kangwook Lee

ArXiv equal contribution

Papers

Is the Importance Ratio Necessary for Stable Reinforcement Learning in LLMs?

Shuibai Zhang*, Junhyuck Kim*, Gyeongman Kim, Jaewoong Cho

ICLR 2026 (SPOT Workshop: Scaling Post-training for LLMs) equal contribution

Planner Aware Path Learning in Diffusion Language Models Training

Fred Zhangzhi Peng, Zachary Bezemek, Jarrid Rector-Brooks, Shuibai Zhang, Anru R. Zhang, Michael Bronstein, Avishek Joey Bose, Alexander Tong

ICLR 2026 Oral Presentation

ParallelBench: Understanding the Tradeoffs of Parallel Decoding in Diffusion LLMs

Wonjun Kang, Kevin Galim, Seunghyuk Oh, Minjae Lee, Yuchen Zeng, Shuibai Zhang, Coleman Hooper, Yuezhou Hu, Hyung Il Koo, Nam Ik Cho, Kangwook Lee

ICLR 2026

dVLM-AD: Enhance Diffusion Vision-Language-Model for Driving via Controllable Reasoning

Yingzi Ma, Yulong Cao, Wenhao Ding, Shuibai Zhang, Yan Wang, Boris Ivanovic, Ming Jiang, Marco Pavone, Chaowei Xiao

CVPR 2026 Findings

VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data

Thomas Zeng, Shuibai Zhang, et al., Kannan Ramchandran, Dimitris Papailiopoulos, Kangwook Lee

ICML 2025 Oral Presentation

Dynamic Multimodal Evaluation with Flexible Complexity by Vision-Language Bootstrapping

Yue Yang*, Shuibai Zhang*, Wenqi Shao*, Kaipeng Zhang, Yi Bin, Yu Wang, Ping Luo

ICLR 2025 Oral Presentation equal contribution

Needle In A Multimodal Haystack

Weiyun Wang*, Shuibo Zhang*, Yiming Ren*, Yuchen Duan*, Tiantong Li*, et al., Wenhai Wang