Segwang Kim

I am a PhD student working on deep learning at Seoul National University.

Welcome to Segwang’s Homepage!

Here are my CV, Github, and MILAB.
I am always welcome to discuss the potentials and limitations of deep learning models’ expressivity.
In particular, I am interested in the compositional generalization abilities of deep learning sequence-to-sequence models.
You can contact me through e-mail (ksk5693@snu.ac.kr) or LinkedIn.

Publications

Conference Proceedings

Segwang Kim, Hyoungwook Nam, Joonyoung Kim, and Kyomin Jung, Neural Sequence-to-grid Module for Learning Symbolic Rules, AAAI Conference on Artificial Intelligence (AAAI) - Feb 2021, A Virtual Conference [code, poster, slides, abridged slides]

Hyoungwook Nam, Segwang Kim, Kyomin Jung, Number Sequence Prediction Problems for Evaluating Computational Powers of Neural Networks, AAAI Conference on Artificial Intelligence (AAAI) - (Oral), Jan 2019, Honolulu, Hawaii, USA [poster, slides]

Journals

Segwang Kim, Joonyoung Kim, and Kyomin Jung, Compositional Generalization via Parsing Tree Annotation, IEEE ACCESS 2021 [code]

Education

PhD, Department of Electrical and Computer Engineering, Seoul National University

Advisor: Kyomin Jung, Mar 2016 -

BA, College of Liberal Studies (major: mathematics), Seoul National University

Mar 2012 - Feb 2016

Materials

Slides

SlideShare

Notes

Coursework-Information Theory
Recap-Probability Theory