loading...

Resume

Home / Resume

About Me

Hi, I am an AI research engineer at the MIT-IBM AI Lab. My current research interests lie in unsupervised representation learning and generative modeling. Particularly, I would like to leverage our insights on human cognition and learning in order to design novel representation learning approaches.

I received my B.A. degree in Computer Science-Statistics from Columbia University in May 2019. During my undergraduate studies, I worked with Professor Jiook Cha, Hod Lipson, and Tal Malkin.

Recently, I have been fortunately been awarded the NSF CSGrad4US Fellowship for my graduate studies.

  • Name: Seungwook Han
  • Age: 27 Years
  • Job: AI Research Engineer
  • Citizenship: South Korea
  • Residence: USA
  • E-mail: swhan [at] mit [dot] edu
Download CV
Experience
Feb 2020 - Present
AI Research Engineer, MIT-IBM AI Lab

Designed a novel two-step generative model for efficient high-dimensional image generation, outperforming the state-of-the-art BigGAN model with a fraction of the compute; Built a novel density ratio estimator leveraging scaled Bregman divergence; Proposed a new framework of equivariances to improve contrastive learning; Currently investigating the role of negatives in contrastive learning and creating a novel method of hard negative mining.

May 2019 - Jan 2020
Data Science Institute Scholar, Columbia Psychiatry Lab

Built a scalable multi-modal deep learning model that uses high performance computing (HPC) to predict socio-cognitive variables like intelligence using neuroimaging data under Professor Jiook Cha.

Jan 2019 - Sept 2019
Research Assistant, Creative Machines Lab

Investigated uncertainty in deep networks with a meta-learning network under Professor Hod Lipson; Proposed a meta-learning model for CV and NLP tasks to use the intermediate layers of the respective task’s base network to predict its correctness.

Aug 2018 - May 2019
Research Assistant, Creative Machines Lab

Investigated uncertainty in deep networks with a meta-learning network under Professor Hod Lipson; Proposed a meta-learning model for CV and NLP tasks to use the intermediate layers of the respective task’s base network to predict its correctness.

Jan 2019 - Sept 2019
Research Assistant, Creative Machines Lab

Investigated uncertainty in deep networks with a meta-learning network under Professor Hod Lipson; Proposed a meta-learning model for CV and NLP tasks to use the intermediate layers of the respective task’s base network to predict its correctness.

Education
Class of 2019
Columbia University

B.A. in Computer Science-Statistics (Data Science). Graduated with magna cum laude.

Class of 2013
Seoul International School

High school diploma.

Skills
  • PyTorch
    95%
  • TensorFlow
    70%
  • AWS
    75%
  • Web Design
    60%
Languages Skills
  • English
    100%
  • Korean
    100%
  • Spanish
    50%