About Me

I’m a fourth year Ph.D student from the College of Information Science & Electronic Engineering, Zhejiang University. My research interests include model compression (typically vector quantization) and edge accelerators. Currently, I’m investigating QAT for large foundation models and inference infra for world-action models.

I’m fortunate to be advised by Prof. Haibin Shen. Currently, I’m a research intern at Robbyant, mentored by Prof. Yinghao Xu and Dr. Zifan Shi. I’m also working closely with Prof. Haotong Qin and Prof. Chaojian Li. I used to be a research intern at VIVO Mobile Communication.

🔥 News

• 2026.09: 🎉 DSAQuant is accepted by NeurIPS 2026.
• 2026.07: 🎉 Lingbot-VA2.0 is released!
• 2026.01: 🎉 ResComp is accepted by ICLR 2026.
• 2025.06: 🎉 SSVQ & ViM-VQ accepted by ICCV 2025.
• 2024.12: 🎉 VQ4DiT is accepted by AAAI 2025.
• 2024.10: 🎉 MVQ is accepted by ASPLOS 2025.

📝 Publications

NeurIPS 2026

DSAQuant: Denoising-Stage-Aligned Quantization-Aware Training for Video Generation

Shuaiting Li, Zelin Gao, Haibin Shen, Yujun Shen, Haotong Qin, Yinghao Xu

• A QAT framework aligned with the stage-wise nature of video diffusion. website code

TECH REPORT

Native Video-Action Pretraining for Generalizable Robot Control

Robbyant Team

• A video-action foundation model built from the ground up for embodiment. website

ICLR2026

Rethinking Residual Errors in Compensation-based LLM Quantization

Shuaiting Li, Juncan Deng, Kedong Xu, Rongtao Deng, Hong Gu, Minghan Jiang, Haibin Shen, Kejie Huang

• A LLM PTQ algorithm, generalization to GPTAQ. code

ASPLOS 2025

MVQ: Towards Efficient DNN Compression and Acceleration with Masked Vector Quantization

Shuaiting Li, Chengxuan Wang, Juncan Deng, Zeyu Wang, Zewen Ye, Zongsheng Wang, Haibin Shen, Kejie Huang

• A highly efficient HW-SW co-designed system for Vector Quantization.

ICCV 2025

SSVQ: Unleashing the Potential of Vector Quantization with Sign-Splitting

Shuaiting Li, Juncan Deng, Chenxuan Wang, Kedong Xu, Rongtao Deng, Hong Gu, Haibin Shen, Kejie Huang

• A new Vector Quantization paradigm, simple but effective. code

ICCV 2025

ViM-VQ: Efficient Post-Training Vector Quantization for Visual Mamba

Juncan Deng, Shuaiting Li, Zeyu Wang, Kedong Xu, Hong Gu, Kejie Huang

• The first sub-2bit Vector Quantization solution for Visual Mambas.

AAAI 2025

VQ4DiT: Efficient Post-Training Vector Quantization for Diffusion Transformers

Juncan Deng, Shuaiting Li, Zeyu Wang, Hong Gu, Kedong Xu, Kejie Huang

• The first 2bit Vector Quantization solution for Diffusion in Transformers. code

• Arxiv Efficiency meets fidelity: A novel quantization framework for stable diffusion, Shuaiting Li, Juncan Deng, Zeyu Wang, Kedong Xu, Rongtao Deng, Hong Gu, Haibin Shen, Kejie Huang
• TCAS-II Ews: An energy-efficient cnn accelerator with enhanced weight stationary dataflow, Chengxuan Wang, Zongsheng Wang, Shuaiting Li, Yuanming Zhang, Haibin Shen, Kejie Huang.
• GRSL Cross-modal adaptation for object detection in thermal infrared remote sensing imagery, Zeyu Wang, Shuaiting Li, Kejie Huang.
• ISCAS A 1FeFET-1T-1C based Compute-in-Memory Macro with Capacitor Reused Pipeline SAR ADC, Minghan Jiang, Rui Xiao, Xinran Li, Shuaiting Li, Xiaolei Zhu, Yishu Zhang, Haibin Shen, Kejie Huang.
• TCASAI RARPIM: A Robust and Adaptive 1FeFET-1T-1C based Compute-in-Memory Macro with Capacitor Reused Pipeline SAR ADC, Minghan Jiang, Rui Xiao, Xinran Li, Shuaiting Li, Xiaolei Zhu, Yishu Zhang, Haibin Shen, Kejie Huang.

📖 Educations

• 2023.09 - now, PhD student, College of information science & electronic engineering, Zhejiang University
• 2019.09 - 2023.06, Undergrad student, College of information science & electronic engineering, Zhejiang University

💻 Internships

• *2024.03 - 2025.9, Research intern at VIVO Mobile Communication, China.