Scholar
Yang Sui
Google Scholar ID: Q2W1p6sAAAAJ
Postdoc, Rice University
Citations & Impact
Citations
941
H-index
13
i10-index
16
Publications
20
Co-authors
15
list available
Publications
Resume (English only)
Academic Achievements
- Publications:
- - September 2025: Two papers accepted by NeurIPS 2025.
- - August 2025: One paper accepted by TMLR 2025.
- - May 2025: One paper accepted by TMLR 2025.
- - April 2025: DFloat11: Lossless Compression for LLM reported by 新智元 and 机器之心.
- - April 2025: Stop Overthinking survey reported by 新智元.
- - March 2025: Released survey: Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models.
- - February 2025: Three papers accepted by CVPR 2025, one of which, TopV, marks my first experience as an advisor.
- - September 2024: One paper accepted by The Thirty-eighth Annual Conference on Neural Information Processing Systems.
Research Experience
- Spring 2024: Research Intern at Snap Research Creative Vision team, proposed 1.99 bits quantization on text-to-image generative model, BitsFusion.
- 2022: Research Intern at Tencent America Media Lab, exploring efficiency and robustness of Learned Image Compression and Transformer models.
- 2019: Full-time Algorithm Engineer at JD, working on face verification and recognition.
- 2018: R&D Intern and member of PaddlePaddle at Baidu, initialized the Paddle-Lite deep learning inference framework.
Education
- Ph.D.: Rutgers University, Advisor: Prof. Bo Yuan.
Background
- Research Interests: Efficient AI and Trustworthy AI. In the domain of Efficient AI, focuses on developing resource-efficient deep learning models without compromising accuracy or performance. In Trustworthy AI, investigates model vulnerability and robustness through adversarial and backdoor attacks.
Miscellany
- Personal Interests: Basketball, DOTA/DOTA2, World of Warcraft. Loves Tracy McGrady, Stephen Curry, Lionel Messi, PIS (YaphetS).
