Research Seminar with Prof. Timothy Baldwin: “Fact-checking Language Models: Generating the Truth, the Whole Truth, and Nothing but the Truth”
We are thrilled to invite you to a thought-provoking research seminar titled:
“Fact-checking Language Models: Generating the Truth, the Whole Truth, and Nothing but the Truth”
Presented by Prof. Timothy Baldwin, Provost of Mohamed Bin Zayed University of Artificial Intelligence.
Details of the Event
- Date: Monday, December 16, 2024
- Time: 10:30 AM – 12:00 PM
- Venue: I201 – Peace Room, VinUni
Tentative Agenda
- 10:30 – 10:35: Welcome and Introduction
- 10:35 – 11:30: Research Seminar
- 11:30 – 12:00: Research/Graduate Opportunities for VinUni students
About the Seminar
Language models (LMs) are widely used in various applications, yet they are notorious for hallucinating or generating outputs that are not grounded in truth.
In this seminar, Prof. Tim Baldwin will explore:
- The challenges of fact-checking textual outputs from language models.
- The causes of hallucinations in LMs and how modified inferencing can reduce them.
- Techniques for improving and evaluating uncertainty quantification in LMs.
This session will offer insights into the future of AI and practical methods for enhancing the reliability of language models.
About the Speaker
Prof. Timothy Baldwin is an esteemed scholar and AI leader, currently serving as the Provost and Professor of Natural Language Processing at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI). He is also a Melbourne Laureate Professor at The University of Melbourne and Chief Scientist of LibrAI, an AI safety start-up.
Academic and Professional Highlights:
- Holds degrees from The University of Melbourne and Tokyo Institute of Technology.
- Authored over 500 peer-reviewed publications in natural language processing and AI.
- Research funded by leading organizations, including Google, Microsoft, ByteDance, and Fujitsu.
- Recipient of multiple awards at top AI conferences and an ARC Future Fellow.
