研讨会信息
🎤 Speaker: Dr. Kailong WANG,
Associate Professor,
School of CSE at Huazhong University of Science and Technology,
(Ph.D. in Computer Science at National University of Singapore)
📰 Title: Towards Trustworthy and Equitable AI in the LLM Era
⏰ Time: 2:30 – 3:30 PM,
📆 Date: Nov. 11, 2025 (Tuesday)
📍 Venue: C2-108, GZ Campus
Online Zoom Meeting:
https://hkust-gz-edu-cn.zoom.us/j/99781303741?pwd=fqOLbfx8MPbCBHHflobKGQzQQoeblH.1
Meeting ID: 997 8130 3741
Passcode: ait
研讨会概要
Abstract
As large language models (LLMs) increasingly underpin applications across education, healthcare, finance, and governance, ensuring their trustworthiness and reliability has become a pressing research frontier. This talk presents a systematic exploration of LLM safety and societal alignment, centered on two representative studies. The first focuses on coverage-guided jailbreak detection, which uses neuron activation patterns to find abnormal model behaviors. This method improves jailbreak detection accuracy, helps prioritize risky test cases, and guides test case generation for robust model evaluation. The second focuses on logic-based hallucination detection, which applies logic reasoning and metamorphic testing to spot factual inconsistencies in LLM outputs. The end-to-end detection framework automatically builds benchmark datasets, verifies answers against trusted knowledge bases, and detects fact-conflicting hallucinations with minimal human effort. Building on these foundations, I will outline future directions towards ensuring LLM trustworthiness and reliability, including developing explainable testing frameworks, conducting empirical studies on LLM bias and its influence on human behavior, and designing formally verified reasoning mechanisms. Together, these efforts aim to advance the Greater Bay Area’s vision for safe, transparent, and human-aligned AI ecosystems.
分享者简介
Dr. Kailong WANG
Associate Professor,
School of Computer Science & Technology,
Dr. Kailong Wang is currently an Associate Professor in the School of CSE at Huazhong University of Science and Technology. He received his Ph.D. in Computer Science from the National University of Singapore and his B.Eng. (First Class Honours) from Nanyang Technological University. His research focuses on trustworthy artificial intelligence, particularly the safety, security, and societal alignment of large language models. Dr. Wang has published extensively in top-tier venues such as OOPSLA, ICSE, FSE, TOSEM, NDSS, ICML and MM. He has received multiple distinctions, including the National Distinguished Postdoctoral Researcher Award and the Outstanding Young Researcher Award (Hubei Province). His broader vision aims to advance safe, transparent, and human-aligned AI to support the Greater Bay Area’s growing innovation ecosystem.
扫描加关注
获取更多AI学域消息
SCAN & FOLLOW US!