Junmo Kang

Ph.D. Student, Georgia Tech

School of Interactive Computing

Coda 1147M

junmo.kang [AT] gatech.edu

About

Hello! I am currently a third-year Ph.D. student in the School of Interactive Computing at Georgia Tech advised by Alan Ritter and Wei Xu. Previously, I completed my M.S. in Computer Science at KAIST, where I started my NLP journey. I had a wonderful time interning at MIT-IBM Watson AI Lab twice.

My research focuses on developing NLP models that are efficient and robust, with the goal of ensuring their practicality in real-world scenarios.

How to induce NLP systems to be more scalable and cheaper in terms of data, compute, or parameters?
How to design NLP models that remain robust when confronted with unseen cases in the wild?

In particular, below are some keywords of my recent interests:

Expert Large Language Models
Instruction-Driven Learning
Retrieval-Augmentation
Efficient NLP
Robustness

News

[May. 2024] I will be returning to MIT-IBM Watson AI Lab as a research intern!
[May. 2024] Self-Specialization paper has been accepted to ACL 2024 Findings!
[Mar. 2024] Our paper on retrieval-augmented LMs' robustness has been accepted to NAACL 2024 Findings.
[Nov. 2023] I passed my PhD Qualifying Exam.
[Oct. 2023] Our new preprint on self-alingment of LLMs for specialization is out.
[May. 2023] I will be spending this summer in Cambridge, MA as a research intern at MIT-IBM Watson AI Lab!
[May. 2023] My first PhD work on cost-efficient fine-tuning of compact models has been accepted to ACL 2023!
[Oct. 2022] Our paper on Graph-Induced Transformers (GIT) has been accepted to EMNLP 2022.
[Aug. 2022] Started my Ph.D. journey at Georgia Tech!

Research Experience

MIT-IBM Watson AI Lab Cambridge, MA, USA

Research Intern May. 2024 - Aug. 2024

Research Intern May. 2023 - Aug. 2023

Host: Leonid Karlinsky, Rogerio Feris

Georgia Institute of Technology Atlanta, GA, USA

Graduate Research Assistant (Ph.D.) Aug. 2022 - Present

Advisor: Alan Ritter, Wei Xu

KAIST IR&NLP Lab Daejeon, Republic of Korea

Research Associate Mar. 2021 - July. 2022

Graduate Research Assistant (M.S.) Feb. 2019 - Feb. 2021

Advisor: Sung-Hyon Myaeng

Publications

^* indicates equal contribution.

Self-MoE: Towards Compositional Large Language Models with Self-Specialized Experts

Junmo Kang, Leonid Karlinsky, Hongyin Luo, Zhen Wang, Jacob Hansen, James Glass, David Cox, Rameswar Panda, Rogerio Feris, Alan Ritter

Preprint

Paper BibTeX

Schema-Driven Information Extraction from Heterogeneous Tables

Fan Bai, Junmo Kang, Gabriel Stanovsky, Dayne Freitag, Mark Dredze, Alan Ritter

Preprint

Paper BibTeX

Self-Specialization: Uncovering Latent Expertise within Large Language Models

Junmo Kang, Hongyin Luo, Yada Zhu, Jacob Hansen, James Glass, David Cox, Alan Ritter, Rogerio Feris, Leonid Karlinsky

ACL 2024 Findings

Paper Poster Blog BibTeX

Why So Gullible? Enhancing the Robustness of Retrieval-Augmented Models against Counterfactual Noise

Giwon Hong*, Jeonghwan Kim*, Junmo Kang*, Sung-Hyon Myaeng, Joyce Jiyoung Whang

NAACL 2024 Findings

Paper BibTeX

Distill or Annotate? Cost-Efficient Fine-Tuning of Compact Models

Junmo Kang, Wei Xu, Alan Ritter

ACL 2023

Paper Slides Poster Code BibTeX

Graph-Induced Transformers for Efficient Multi-Hop Question Answering

Giwon Hong, Jeonghwan Kim, Junmo Kang, Sung-Hyon Myaeng

EMNLP 2022

Paper BibTeX

Exploiting Numerical-Contextual Knowledge to Improve Numerical Reasoning in Question Answering

Jeonghwan Kim, Junmo Kang, Giwon Hong, Kyung-min Kim, Sung-Hyon Myaeng

NAACL 2022 Findings

Paper BibTeX

Ultra-High Dimensional Sparse Representations with Binarization for Efficient Text Retrieval

Kyoung-Rok Jang, Junmo Kang, Giwon Hong, Sung-Hyon Myaeng, Joohee Park, Taewon Yoon, Heecheol Seo

EMNLP 2021

Paper BibTeX

Leveraging Order-Free Tag Relations for Context-Aware Recommendation

Junmo Kang, Jeonghwan Kim, Suwon Shin, Sung-Hyon Myaeng

EMNLP 2021

Paper Poster BibTeX

Have You Seen That Number? Investigating Extrapolation in Question Answering Models

Jeonghwan Kim, Giwon Hong, Kyung-min Kim, Junmo Kang, Sung-Hyon Myaeng

EMNLP 2021

Paper Poster BibTeX

Can You Distinguish Truthful from Fake Reviews? User Analysis and Assistance Tool for Fake Review Detection

Jeonghwan Kim*, Junmo Kang*, Suwon Shin*, Sung-Hyon Myaeng

HCI+NLP@EACL 2021

Paper Poster BibTeX

Regularization of Distinct Strategies for Unsupervised Question Generation

Junmo Kang*, Giwon Hong*, Haritz Puerto San Roman*, Sung-Hyon Myaeng

EMNLP 2020 Findings

Paper BibTeX

Handling Anomalies of Synthetic Questions in Unsupervised Question Answering

Giwon Hong*, Junmo Kang*, Doyeon Lim*, Sung-Hyon Myaeng

COLING 2020

Paper BibTeX

Let Me Know What to Ask: Interrogative-Word-Aware Question Generation

Junmo Kang*, Haritz Puerto San Roman*, Sung-Hyon Myaeng

MRQA@EMNLP 2019

Paper Poster BibTeX

Vitæ

Full CV in PDF.

MIT-IBM Watson AI Lab May. 2024 - Aug. 2024

Research Intern
Compositional LLMs of Self-Specialized Experts
MIT-IBM Watson AI Lab May. 2023 - Aug. 2023

Research Intern
Self-Alignment of LLMs for Specialization
Georgia Tech Aug. 2022 - Present

Ph.D. in Computer Science
KAIST IR&NLP Lab Mar. 2021 - July. 2022

Research Associate
QA & IR
KAIST Feb. 2019 - Feb. 2021

M.S. in Computer Science
IR&NLP Lab
Republic of Korea Army Apr. 2013 - Jan. 2015

Honorably discharged as Sergeant
Compulsory military service
Chungnam National University Mar. 2012 - Feb. 2019

B.E. in Computer Science & Engineering
Summa Cum Laude