Hi! I am Shigeng Wang (王世耿), a Ph.D. student in Computer Science at Beijing University of Posts and Telecommunications (BUPT), supervised by Zhonghong Ou. I obtained my Bachelor’s degree from BUPT and continued my Ph.D. studies at BUPT through the direct doctoral track, with an expected graduation in June 2026.

I am currently conducting research at Intel Labs China as an AI Research Intern, supervised by Anbang Yao. My research focuses on large language model quantization, compression, and hardware-efficient acceleration. More broadly, my interests lie in computer vision and efficient deep learning, with particular emphasis on model quantization and lightweight inference.

🔥 News

• 2026.01: A first-author paper (SliderQuant) on post-training quantization for large language models was accepted to ICLR 2026.
• 2026.01: Three first-author papers were accepted to ICASSP 2026.
• 2025.11: The Yingtao App, an AI PC intelligent assistant powered by SliderQuant, was released by Intel, with me as a core developer.
• 2023.01: A co-authored paper on machine learning for fungal keratitis diagnosis was accepted to eBioMedicine.
• 2022.11: Contributed to the development of ppq_tools, a user-friendly model quantization toolkit.
• 2021.10: A first-author paper on lightweight object detection was accepted to CCIS 2021.

📝 Publications

(* Equal contribution, # Corresponding author)

Selected Publications

ICLR 2026

SliderQuant: Accurate Post-Training Quantization for Large Language Models

Shigeng Wang*, Chao Li*, Yangyuxuan Kang, Jiawei Fan, Zhonghong Ou, Anbang Yao#.

Code | Project (to be released)

SliderQuant is a post-training quantization framework that adaptively accounts for layer-wise sensitivity in large language models, achieving strong performance gains under ultra-low-bit weight and weight-activation quantization.

• Dynabits: Token-Aware Weight–Activation Quantization for Large Vision–Language Models, Shigeng Wang, Zhonghong Ou#, Hongxing Zhang, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2026).
• VARDet: Visual Autoregressive Multi-Scale Prediction and CLIP-Guided Semantics for UAV Small-Object Detection, Shigeng Wang, Zhonghong Ou#, Hongxing Zhang, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2026).
• Foreground-Enhanced Coarse-to-Fine Detection for UAV Small Objects, Shigeng Wang, Zhonghong Ou#, Hongxing Zhang, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2026).

• Development and Multi-Center Validation of a Machine Learning Model for Early Detection of Fungal Keratitis, Zhenyu Wei, Shigeng Wang, Zhiqun Wang, et al., eBioMedicine, 2023.

• PowerDet: Efficient and Lightweight Object Detection for Electric Power Open Scenes, Shigeng Wang, Zhonghong Ou#, Meina Song, IEEE 7th International Conference on Cloud Computing and Intelligent Systems (CCIS 2021).

📖 Educations

• 2021.09 – 2026.06, Ph.D. student in Computer Science, Beijing University of Posts and Telecommunications (BUPT).

• 2017.09 – 2021.06, B.Eng. in Data Science and Big Data Technology, Beijing University of Posts and Telecommunications (BUPT).

🎖 Honors and Awards

• 2022.10: First-Class Scholarship, Beijing University of Posts and Telecommunications.
• 2022.06: BUPT Excellent Ph.D. Students Foundation.
• 2021.06: Outstanding Graduate of Beijing.
• 2020.09: First Prize, National Undergraduate Mathematics Competition.
• 2020.07: First Prize, China National Undergraduate E-Commerce Innovation, Creativity and Entrepreneurship Challenge.
• 2019.09: Second Prize, China Undergraduate Mathematical Modeling Competition.
• 2018.10: National Encouragement Scholarship (China).

💬 Academic service

Reviewer: ICML 2026, PRCV 2025.

💻 Experieces

• 2024.04 - now, Intel Labs, China. Research Intern, supervised by Anbang Yao.
• 2023.10 – 2024.03, QCraft, China. Research Intern, focusing on autonomous driving perception.