Jiawei LIU

semiahmoo.jpeg

I am a CS PhD student at UIUC advised by Lingming Zhang since 2021 Fall.

My primary research goal is to make future software infrastructures: easy-to-use, high-performance and reliable. At present, I am developing PLSE techniques to make ML Systems reliable and efficient.

ūü§ó Open to collaborate! We should chat if you are also interested in fuzzing and building ML Systems!

News

Jan 25, 2022 I will join OctoML as a Summer intern to build the next-generation TVM!
Dec 16, 2021 Our tensor compiler fuzzing work, Tzer, accepted by OOPSLA’22!

Publications

  1. Pre-print
    Finding Deep-Learning Compilation Bugs with NNSmith
    Jiawei Liu, Jinkun Lin, Fabian Ruffy, Cheng Tan, Jinyang Li, Aurojit Panda,  and Lingming Zhang
    arXiv preprint arXiv:2207.13066 2022
  2. OOPSLA’22Atifact AvailableAtifact Reusable
    Coverage-guided tensor compiler fuzzing with joint IR-pass mutation
    Jiawei Liu, Yuxiang Wei, Sen Yang, Yinlin Deng,  and Lingming Zhang
    Proceedings of the ACM on Programming Languages 6 (OOPSLA1), Apr 2022
  3. ACMMM’21 (OSC)
    Fast and Flexible Human Pose Estimation with HyperPose
    Yixiao Guo*,  Jiawei Liu*, Guo Li*, Luo Mai,  and Hao Dong
    Proceedings of the 29th ACM International Conference on Multimedia, Apr 2021

Internship

OctoML, Smr. 2022   Pattern Language

Alibaba DAMO, Smr. 2021 GNN Serving

Systems@NYU, Smr.+Fall 2020 Video Analytics

ByteDance AI Lab, Spr. 2020   DNN Serving

Services

Artifact Evaluation Committee: OSDI'22, ATC'22

Talks

Coverage-Guided Tensor Compiler Fuzzing with Joint IR-Pass Mutation
  • SAMPL Lunch Talks, University of Washington [Video] May 2022
  • Software Engineering Seminar, University of Illinois at Urbana-Champaign April 2022