TUTORIALS

Towards Robust and Trustworthy Large Language Models: Issues and Mitigation Strategies

Speakers

Cheng-Kuang Wu (Appier brian.wu@appier.com), Zhi Rui Tam (Appier ray.tam@appier.com), and Kuan-Hao Huang (Texas A&M University khhuang@tamu.edu)

Abstract

Large language models (LLMs) have recently demonstrated strong capabilities across numerous downstream tasks, significantly expanding their potential for real-world applications. However, their deployment remains limited by critical robustness and trustworthiness challenges. This tutorial provides attendees with a comprehensive overview of these critical issues, their corresponding mitigation strategies, and open research questions in this rapidly evolving area.

Machine Learning for Natural Science: breakthrough in the protein universe and other scientific areas

Speakers

Joshua Yao-Yu Lin (Genentech joshualin24@gmail.com)

Abstract

Machine learning is transforming how the world operates—can it also reshape natural science? In this talk, I will highlight some of the exciting works, starting from biology: tackling the grand challenge of protein structure prediction with AlphaFold I, II, and III, and moving to generative models for de novo protein design such as RFdiffusion and inverse folding models like ProteinMPNN, along with their potential applications and challenges for drug design. I will then touch on other fields, including mathematics, with development like AlphaEvolve for extending the boundaries of math and AI, as well as new models that tackle IMO-level problems, and the growing role of AI as a co-scientist in guiding experiments. Finally, I will mention applications in physics and astronomy, from exoplanet discovery to mapping cosmic structures, showing how AI is becoming an emerging collaborator in advancing fundamental science.