Qidong Su 蘇起冬


Update: Checkout out APPL, a new language designed for writing prompts for LLMs in a more structured and efficient way.

About me

I am a Computer Science PhD student in University of Toronto, advised by Gennady Pekhimenko. I am a member of Computer System and Networks group and EcoSystem group. Before that, I got my bachelor's degree in Computer Science from Shanghai Jiao Tong University, where I was a member of ACM honors class.

My name is pronounced as /tɕʰi tʊŋ/.

Research

My research revolves around the evolution of the software stack in the new era of AI, especially about its programmability. I'm always curious about the how the computer system stack works as a whole (how abstract programs can run on hardwares made of transistors), and how we can improve its efficiency (by new hardwares/softwares, compilers or programming models/languages). Some topics I am interested in:

  • Large language models
  • Compilers, code generation
  • Irregular data processing
  • Robustness of ML systems and algorithms

I was a research intern of Amazon Shanghai AI Lab, where I worked on a distributed graph neural network framework based on Deep Graph Library (DGL). I was fortunate to be mentored by Dr. Minjie Wang.

Publications

Service

  • MLSys'23 Artifact Evaluation

Misc

Language: Chinese(Mandarin, Wu-native, Cantonese-fluent), English(fluent), Japanese(beginner)

An amateur in linguistics (phonetics, Chinese dialects). Fan of City Pop and Enka.

Contact

  • Email: qdsu ät cs.toronto.edu
Last modified by Qidong Su. (Jun.26 2024)