profile photo

Jihoon Tack


Contact: jihoontack@{kaist.ac.kr, gmail.com}

About

I am a final year Ph.D. student at KAIST, advised by Jinwoo Shin and a Research Intern at Meta FAIR, advised by Xian Li. I am also closely collaborating with Jonathan Richard Schwarz at Harvard University and Yee Whye Teh at University of Oxford. My research is primarily centered on developing efficient and robust (or safe) machine learning frameworks to tackle the emerging challenges of large models. In particular, my recent research focuses on developing efficient pre-training and adaptation algorithms, leveraging useful prior knowledge extracted from multiple tasks (e.g., meta-learning) combined with efficient training and inference schemes (e.g., data pruning, weight pruning, quantization, and efficient parametrization). I am fortunate to be a recipient of Google Ph.D. Fellowship in machine learning.

Topics of interest (publications with overlap):

Publications

* denotes equal contribution

Sparsified State-Space Models are Efficient Highway Networks
Woomin Song, Jihoon Tack, Sangwoo Mo, Seunghyuk Oh, Jinwoo Shin
NeurIPS Workshop on Efficient Natural Language and Speech Processing 2024, Oral presentation
paper | code
Online Adaptation of Language Models with a Memory of Amortized Contexts
Jihoon Tack, Jaehyung Kim, Eric Mitchell, Jinwoo Shin, Yee Whye Teh, Jonathan Richard Schwarz
NeurIPS 2024
paper | code | slide | project page
ReMoDetect: Reward Models Recognize Aligned LLM's Generations
Hyunseok Lee*, Jihoon Tack*, Jinwoo Shin
NeurIPS 2024
paper | code | project page
Optimized Feature Generation for Tabular Data via LLMs with Decision Tree Reasoning
Jaehyun Nam*, Kyuyoung Kim*, Seunghyuk Oh, Jihoon Tack, Jaehyung Kim, Jinwoo Shin
NeurIPS 2024
paper | code
Tabular Transfer Learning via Prompting LLMs
Jaehyun Nam, Woomin Song, Seong Hyeon Park, Jihoon Tack, Sukmin Yun, Jaehyung Kim, Jinwoo Shin
COLM 2024
paper | code
Unleashing the Power of Meta-tuning for Few-shot Generalization Through Sparse Interpolated Experts
Shengzhuang Chen, Jihoon Tack, Yunqiao Yang, Yee Whye Teh, Jonathan Richard Schwarz, Ying Wei
ICML 2024
paper | code
Learning Large-scale Neural Fields via Context Pruned Meta-Learning
Jihoon Tack, Subin Kim, Sihyun Yu, Jaeho Lee, Jinwoo Shin, Jonathan Richard Schwarz
NeurIPS 2023
paper | code | slide | poster
Modality-Agnostic Self-Supervised Learning with Meta-Learned Masked Auto-Encoder
Huiwon Jang*, Jihoon Tack*, Daewon Choi, Jongheon Jeong, Jinwoo Shin
NeurIPS 2023
paper | code
Modality-Agnostic Variational Compression of Implicit Neural Representations
Jihoon Tack*, Jonathan Richard Schwarz*, Yee Whye Teh, Jaeho Lee, Jinwoo Shin
ICML 2023
paper | poster
STUNT: Few-shot Tabular Learning with Self-generated Tasks from Unlabeled Tables
Jaehyun Nam, Jihoon Tack, Kyungmin Lee, Hankook Lee, Jinwoo Shin
ICLR 2023, Spotlight presentation
Samsung Humantech Paper Awards 2023, Bronze Prize
paper | code
Rethinking the Entropy of Instance in Adversarial Training
Minseon Kim, Jihoon Tack, Jinwoo Shin, Sung Ju Hwang
IEEE SaTML 2023
paper
Meta-Learning with Self-Improving Momentum Target
Jihoon Tack, Jongjin Park, Hankook Lee, Jaeho Lee, Jinwoo Shin
NeurIPS 2022
paper | code | slide | poster
K-centered Patch Sampling for Efficient Video Recognition
Seong Hyeon Park, Jihoon Tack, Byeongho Heo, Jung-Woo Ha, Jinwoo Shin
ECCV 2022
paper | code
Generating Videos with Dynamics-aware Implicit Generative Adversarial Networks
Sihyun Yu*, Jihoon Tack*, Sangwoo Mo*, Hyunsu Kim, Junho Kim, Jung-Woo Ha, Jinwoo Shin
ICLR 2022
paper | code | slide | project page
Consistency Regularization for Adversarial Robustness
Jihoon Tack, Sihyun Yu, Jongheon Jeong, Minseon Kim, Sung Ju Hwang, Jinwoo Shin
AAAI 2022
ICML Workshop on Adversarial Machine Learning 2021, Oral presentation
Best Paper Award, Korean Artificial Intelligence Association 2021
paper | code | slide | poster
Meta-Learning Sparse Implicit Neural Representations
Jihoon Tack*, Jaeho Lee*, Namhoon Lee, Jinwoo Shin
NeurIPS 2021
paper | code | slide | poster
CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances
Jihoon Tack*, Sangwoo Mo*, Jongheon Jeong, Jinwoo Shin
NeurIPS 2020
Qualcomm Innovation Fellowship Korea 2020
paper | code | slide | poster
Adversarial Self-Supervised Contrastive Learning
Minseon Kim, Jihoon Tack, Sung Ju Hwang
NeurIPS 2020
paper | code | project page

Education

M.S. & Ph.D. in Artificial Intelligence
Korea Advanced Institute of Science and Technology (KAIST), Mar. 2019 - Present
Advisor: Jinwoo Shin
B.S. in Electrical Engineering and Mathematics (minored)
Korea Advanced Institute of Science and Technology (KAIST), Mar. 2014 - Feb. 2019

Experience

Research Scientist Intern, Meta FAIR (Host: Xian Li), Jun. 2024 - Present
Visiting Student / Collaborator, University of Oxford (Host: Yee Whye Teh), Jun. 2023 - May 2024
External Collaborator, Jonathan Richard Schwarz at Harvard University, Jun. 2022 - May 2024
External Collaborator, Jaeho Lee at POSTECH, Apr. 2022 - May 2023

Honors and Awards

Google Ph.D. Fellowship 2023, Research Area: Machine Learning
Samsung Humantech Paper Awards 2023, Bronze Prize
Best Paper Award, Korean Artificial Intelligence Association 2021
Qualcomm Innovation Fellowship Korea 2020
Conference Scholar Award: NeurIPS 2023
Reviewer Awards: NeurIPS 2020, CVPR 2021, NeurIPS 2022

Academic Activities

Conference Reviewer: NeurIPS, ICML, ICLR, CVPR, ICCV, AAAI
Journal Reviewer: TMLR, IJCV, IEEE TPAMI, IEEE TNNLS, IEEE TIFS, IEEE TIP
Workshop Reviewer: AI4CC@CVPR, VAND@CVPR, Neural-Field@ICLR, DistShift@NeurIPS