Jianchang Su — Ph.D. Candidate, University of Connecticut
About Me
"道阻且长，行则将至"
— 《荀子·修身》
I am a Ph.D. candidate at University of Connecticut in the School of Computing. My research focuses on building efficient, scalable, and trustworthy infrastructure for cloud computing and ML/LLM systems. I work on systems-level optimization for distributed training and serving, with the goal of addressing performance bottlenecks and security challenges in modern computing environments. I received my Master's degree in Artificial Intelligence from Friedrich-Alexander-Universität Erlangen-Nürnberg.
SAIR: Cost-Efficient Multi-Stage ML Pipeline Autoscaling via In-Context Reinforcement Learning accepted to NeurIPS 2026!
Sep 24, 2026
Safe and Programmable OS-Level GPU Resource Management with eBPF accepted to ASPLOS '27! Congratulations to Yusheng Zheng, who led this work, and to the rest of the team.
Aug 20, 2026
Structural Collapse of Safety-Critical Weight Protection Scorers Under Group-Wise Post-Training Quantization accepted to EMNLP 2026 Findings!
gpu_ext: Extensible OS Policies for GPUs via eBPF 4
Zheng, Yusheng and Yu, Tong and Yang, Yiwei and Jiang, Minghui and Gao, Xiangyu and Su, Jianchang and Hu, Yanpeng and Mao, Wenan and Zhang, Wei and Williams, Dan and others
Selective KV-Cache Sharing to Mitigate Timing Side-Channels in LLM Inference 24
Chu, Kexin and Lin, Zecheng and Xiang, Dawei and Shen, Zixu and Su, Jianchang and Chu, Cheng and Yang, Yiwei and Zhang, Wenhui and Wu, Wenfei and Zhang, Wei
eInfer: Unlocking Fine-Grained Tracing for Distributed LLM Inference with eBPF 7
Chu, Kexin and Su, Jianchang and Zhang, Yifan and Zhao, Chenxingyu and Yang, Yiwei and Zheng, Yusheng and Lin, Shengkai and Zhao, Shizhen and Zhang, Wei
Proceedings of the 3rd Workshop on eBPF and Kernel Extensions, 2025 workshop
