Yaohui Cai

I am Yaohui Cai (蔡耀辉). I am now working at Jump Trading as a quantitative researcher.

I earned my Ph.D. from Cornell University advised by Zhiru Zhang and Chris De Sa. I received my bachelor degree from Peking University. My reserach mainly focuses on machine learning for system design and efficient machine learning.

Email: yc2632 [at] cornell [dot] edu

Google Scholar / GitHub / LinkedIn

Selected Publications

HeuriGym: An Agentic Benchmark for LLM-Crafted Heuristics in Combinatorial Optimization

Hongzheng Chen* , Yingheng Wang*, Yaohui Cai*, Hins Hu*, Jiajie Li* , Shirley Huang, Chenhui Deng, Rongjian Liang, Shufeng Kong, Haoxing Ren, Samitha Samaranayake, Carla P. Gomes, Zhiru Zhang.

[paper] [code] International Conference on Learning Representations (ICLR), 2026.

SmoothE: Differentiable E-Graph Extraction

Yaohui Cai, Kaixin Yang, Chenhui Deng, Cunxi Yu, Zhiru Zhang.

[paper] [code] International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), 2025.

(ASPLOS 2025 Best Paper Award)

A Comprehensive Evaluation of FPGA-Based Spatial Acceleration of LLMs

Hongzheng Chen, Niansong Zhang, Yixiao Du, Zichao Yue, Yaohui Cai, Zhiru Zhang.

[paper] ACM Transactions on Reconfigurable Technology and Systems (TRETS), 2024.

QuIP: 2-Bit Quantization of Large Language Models With Guarantees

Jerry Chee, Yaohui Cai, Volodymyr Kuleshov, Chris De Sa.

[paper] [code] Conference on Neural Information Processing Systems (NeurIPS), 2023.

(Spotlight)

ZeroQ: A Novel Zero Shot Quantization Framework.

Yaohui Cai*, Zhewei Yao*, Zhen Dong*, Amir Gholami, Michael Mahoney, and Kurt Keutzer.

[paper] [code] Conference on Computer Vision and Pattern Recognition (CVPR), 2020.

Experiences

Google Research, Brain Team (Now Google DeepMind)
Student Researcher • May 2023 - Feb. 2024

Meta, Meta Reality Labs
Research Scientist Intern • May 2022 - Dec. 2022

University of California, Berkeley, Berkeley Artificial Intelligence Research (BAIR) Lab
Visiting Student Researcher • May 2019 - Jan. 2020
Advisors: Professor Kurt Keutzer and Professor Michael Mahoney

Website source from Jon Barron and Xingyuan Sun.