Nice to meet you! I am Kathryn (Jinqi) Chen, a senior undergraduate student at Carnegie Mellon University School of Computer Science. I am part of the Catalyst Group advised by Prof. Tianqi Chen working on deep learning compilers. Previously, I was an intern at the MLSys team of OctoML in summer 2022, and contributed to the ApacheTVM project. In the past year, I also worked closely with Huan Zhang on formal verification of large neural networks. Our network verifier α,β-CROWN won the third International Verification of Neural Networks Competition (VNN-COMP'22).
I am interested in understanding the fundamental aspects of deep learning. Currently, I am attracted to the following two areas:
- the system-level characteristics of modern deep learning
- the interpretability of deep learning algorithms from a neuron perspective
The core question I have been thinking about is: what results in the limited capacity of modern deep learning and how could we improve it?
Stay tuned to my publications! I am always willing to chat about new ideas, so feel free to reach out to me at jinqic AT cs DOT cmu DOT edu.
I am hugely interested in learning to make music and DJing. I draw insights from house music (melodic house, deep house), J-pop, rock (alternative rock, progressive rock, industrial rock, post-rock), neo soul and many others. If you have similar tastes in music, I can’t wait to getting to know you!