Department of Industrial Engineering & Decision Analytics [Joint IEDA/ISOM] seminar - How Operations Research Enhances LLM Research
Supporting the below United Nations Sustainable Development Goals:支持以下聯合國可持續發展目標：支持以下联合国可持续发展目标：
This talk explores how optimization techniques help the research of large language models. The first part demonstrates how optimization techniques address challenges in post-training alignment of LLMs. We introduce a novel preference alignment method using comparison oracles with proven convergence guarantees, which effectively mitigates the likelihood displacement issues found in traditional direct alignment approaches. Our method shows particular strength when handling noisy preference pairs with similar likelihoods. Experimental evaluations across multiple models (Mistral-7B, Llama-3-8B, Gemma-2-9B) demonstrate significant improvements compared to existing techniques.
The second part examines game theory applications in AI verification through Proof-of-Learning (PoL), a mechanism that simultaneously addresses blockchain sustainability concerns and AI trustworthiness challenges. We present an incentive-secure PoL system with computational efficiency, controllable difficulty, and robust security guarantees. Our approach includes a novel peer-prediction solution to the "Verifier's Dilemma," establishing theoretical foundations for sustainable blockchain systems, while creating a decentralized platform for AI verification.
Xi Chen is a Professor and Andre-Meyer Faculty Fellow at Stern School of Business, New York University, who is also an affiliated professor at Computer Science and Center for Data Science. Before that, he was a Postdoc in the group of Prof. Michael Jordan at UC Berkeley and obtained his Ph.D. from the Machine Learning Department at Carnegie Mellon University.
He studies high-dimensional machine learning, online learning, large-scale stochastic optimization, and applications to operations management and FinTech. Recently, he started a new research line on blockchain technology and decentralized finance. He is an ASA Fellow, IMS Fellow, recipient of COPSS Leadership Award, NSF Career Award, The World’s Best 40 under 40 MBA Professor by Poets & Quants, and Forbes 30 under 30 in Science.