<2026 Spring Colloquium #2: AI Beyond LLMs: The Potential of Foundation Models for Numerical Data (Prof. Jaemin Yoo)>

ㅇ Speaker: Prof. Jaemin Yoo (Department of Computer Science and Engineering)

ㅇ Date & Time: Thursday, May 14, 2026, 5:00 PM

ㅇ Location: Room 105, Building 302

ㅇ Lecture Overview

    The emergence of large language models has significantly reshaped the landscape of AI technology. A single model capable of addressing a wide range of problems — the “foundation model” — has become the new standard in text and image domains.

However, most of the data we encounter is neither language nor images.

Tabular data made up of numbers and fields — hospital records, financial transactions, process sensor data, customer information — continues to play a critical role in real-world industry and research.

This lecture introduces “numerical data foundation models,” a new approach that goes beyond these limitations.

The core idea behind these models is to learn, during pretraining on millions of synthetic datasets, the very process of “learning from new data.”

As a result, such models open up the possibility of understanding and analyzing new tabular data immediately, without additional training.

ㅇ Speaker Bio
    Professor Jaemin Yoo conducts research on data-driven AI in the SNU Department of Computer Science and Engineering, with a particular interest in foundation models for numerical and structured data.

He worked as a postdoctoral researcher at Carnegie Mellon University and has published numerous papers at leading machine learning and data mining conferences, including NeurIPS, ICML, ICLR, KDD, and WWW.

ㅇ Watch the lecture: https://www.youtube.com/watch?v=DQEds9uxXsc&t=120s