EE · Fall 2026 · Graduate Seminar

Advanced Deep Learning

Contemporary representation learning, foundation models, and physical intelligence through research papers and implementation.

Course Overview

This graduate seminar examines recent advances in deep learning and foundation models. Lectures establish the core ideas behind each topic, while student presentations and discussions critically analyze the contributions, methodology, experiments, and limitations of current research. A semester project develops the ability to reproduce, evaluate, and extend a recent paper.

Instructor
Prof. Nuri Kim
Term
Fall 2026
Meetings
Mon & Wed
11:00–12:15
Classroom
Engineering Building 7
Room 114

Logistics

Learning goals

  • Understand current research directions in deep learning and foundation models.
  • Critically analyze the ideas, methodology, experiments, and limitations of research papers.
  • Reproduce a recent method and propose a well-motivated extension.
  • Communicate technical arguments clearly through presentation and discussion.

Course information

  • Office hours: Monday, 14:00–15:00, Room 326.
  • Language: Korean lectures with English materials.

Recommended background: deep learning, machine learning, linear algebra, probability, algorithms, and optimization.

Schedule

Part I
Learning Paradigms
W01Course IntroductionTBA
W02Self-Supervised LearningTBA
W03Transfer and Continual LearningTBA
W04Meta-LearningTBA
Part II
Generative & Language Models
W05Generative Models ITBA
W06Generative Models IITBA
W07Recent Neural Architectures for LanguageTBA
W08Midterm Period—
W09Large Language ModelsTBA
W10Applications of Large Language ModelsTBA
Part III
Multimodal & Physical AI
W11Vision-Language Foundation ModelsTBA
W12Applications of Vision-Language Foundation ModelsTBA
W13Vision-Language-Action ModelsTBA
W14World ModelsTBA
W15Final Period—

The schedule is tentative. Additional topics may be selected for paper presentations or semester projects.

Assessment

Attendance & Participation

Attend at least 70% of class meetings and participate consistently in lectures and discussions.

Paper Presentation

Give a 15–20 minute talk covering the paper’s motivation, problem, contributions, method, and experiments.

Paper Report

Reproduce results, identify weaknesses, and propose an improvement supported by experiments.

Presentation
20%
Report
60%
Attendance
20%

Generative AI policy: AI tools may be used only within the instructor’s stated scope. Any use must be reviewed, revised, and clearly disclosed in submitted work.

Resources