I graduated with a PhD degree in computer sciences in 2021 from University of Wisconsin-Madison.
I am fortunate to be advised by Prof. Somesh Jha
and Prof. Aws Albarghouthi.
My research interest lies in the intersection of
machine learning, programming language and natural language.
I am specifically interested in big code and naturalness tasks
and word embedding methods.
(2021 June) Notice: This page is no longer actively maintained and may become unavailable in the future.
Check my new personal page for more up-to-date information.
Research Topics
- Representation Learning for Source Code. Leveraging large corpus and naturalness, build (deep) learning models for code
understanding, generation, language modeling and embedding.
- Subword Embedding Methods. Look into word spellings to utilize lexical information and mitigate problems caused by a fixed-size vocabulary.
- Learning from Structures. Towards effective and efficient deep learning over structured data, for example, trees and DAGs.
Papers
- Code Prediction by Feeding Trees to Transformers,
Seohyun Kim*, Jinman Zhao*, Yuchi Tian, Satish Chandra.
(*Equal contribution)
ICSE 2021
/ arXiv
- PBoS: Probabilistic Bag-of-Subwords for Generalizing Word Embedding,
Zhao Jinman, Shawn Zhong, Xiaomin Zhang, Yingyu Liang.
Findings of EMNLP 2020
/ arXiv
/ slides
/ talk
- Generalizing Word Embeddings using Bag of Subwords,
Jinman Zhao, Sidharth Mudgal, Yingyu Liang.
EMNLP 2018
/ arXiv
/ slides
/ talk
- Neural-Augmented Static Analysis of Android Communication,
Jinman Zhao, Aws Albarghouthi, Vaibhav Rastogi, Somesh Jha, Damien Octeau.
FSE 2018
/ arXiv
/ slides
/ poster
- The Effect of Network Width on the Performance of Large-batch Training,
Lingjiao Chen, Hongyi Wang, Jinman Zhao, Dimitris Papailiopoulos, Paraschos Koutris.
NeurIPS 2018
/ arXiv
Selected Reports
- Investigating the skip-gram word embedding model, 2017 Spring.
- Low resolution facial landmark detection, 2016 Spring.
- Learning comment topics from code, 2016 Spring.
Service & Activities
Adapted from
Vitae.
Last updated 2020/11/02.