Seok Min Lim

IMDA

Seok Min leads the technical research team in Infocomm Media Development (IMDA), driving initiatives in trust-related technologies, such as AI testing and digital watermarking.

She began her career in cybersecurity research and has more recently focused on AI governance testing (AI Verify) and the evaluation of generative models (Project Moonshot).

Talk

Title: From Research to Implementation: Testing and Evaluating Large Language Models

Abstract: In today’s rapidly evolving AI landscape, large language models (LLMs) have emerged as a cornerstone of many AI-driven solutions like chatbots and code generation. However, this advancement also introduces a unique set of security and safety challenges, ranging from data privacy risks, biases in model outputs to the risks of generating and executing malicious codes when using these new AI systems. In this talk, we will describe these risks and present the testing approaches that we have included in Project Moonshot for our users to evaluate their models.