AI Visual
Session 1
AI Visual

Foundations of Next-Generation AI

13:30-15:35 (125')
ChairEunho Yang (KAIST)
Hanra Hall 1·2
13:30-13:45 (15')
Ichiro Takeuchi
RIKEN
Research Design in Data-Driven Science: Building a Framework for Reliability & Transparency
13:45-14:00 (15')
Tae-Eui Kam
Korea University
AI-Powered Drug Discovery: From Molecular Property Prediction to Generation
14:00-14:15 (15')
Seong Jae Hwang
Yonsei University
Dissecting Vision-Language Models for Training-Free Solutions
14:15-14:30 (15')
Noseong Park
KAIST
Scientific Foundation Models
14:30-14:45 (15')
Jaeho Lee
POSTECH
Overcoming the Compression Scaling Laws
14:45-15:00 (15')
Karteek Alahari
INRIA
Are Low-Resource Vision-Language Models Viable?
Panel Discussion : Sustainable AI Research Ecosystems in the Era of High-Cost Infrastructure
15:00-15:35 (35')
ChairEunho Yang (KAIST)
Description
This panel discussion explores the challenges and opportunities in advancing AI research amid rising computational costs. Experts will share recent work on cost-efficient AI training, discuss how academia and collaborative communities can thrive despite limited resources, and examine the real-world impact of AI in sectors like healthcare and finance. Join us for insights on building sustainable AI research ecosystems and driving societal innovation.
Panelists
Ichiro Takeuchi
RIKEN
Tae-Eui Kam
Korea University
Seong Jae Hwang
Yonsei University
Noseong Park
KAIST
Jaeho Lee
POSTECH
Karteek Alahari
INRIA
Sangdoo Yun
NAVER Cloud

Principles for New AI Algorithms

15:55-17:50 (115')
ChairJinwoo Shin (KAIST)
Hanra Hall 1·2
15:55-16:15 (20')
Mengye Ren
NYU
Lifelong Concept Learning
16:15-16:35 (20')
Seon Joo Kim
Yonsei University
Understanding Streaming Videos
16:35-16:55 (20')
Julia Stoyanovich
NYU
Responsible Data Engineering
16:55-17:15 (20')
Jaesik Choi
KAIST
Explainable AI to Analyze Internal Decision Mechanism of Deep Neural Networks
Panel Discussion : Foundational Principles Beyond the Current AI Paradigm
17:15-17:50 (35')
ChairJuho Lee (KAIST)
Description
Recent advances in multi-modal foundation models have profoundly transformed the landscape of AI research. While some believe these developments bring us closer to artificial general intelligence (AGI), others remain skeptical. This panel will examine what may still be lacking in current paradigms—for example, how we might design AI systems capable of learning continually and efficiently from real-world interactions, as humans naturally do.
Panelists
Mengye Ren
NYU
Seon Joo Kim
Yonsei University
Julia Stoyanovich
NYU
Jaesik Choi
KAIST
He He
NYU
Rajesh Ranganath
NYU
Steven Whang
KAIST