报告人：Linqi Song, Department of Computer Science, City University of Hong Kong
时间：2026年9月16日下午3:00-5:00，地点：7502
报告题目：Title: Enhancing the Reasoning Abilities of Large Language Models (LLMs)
Abstract: While modern LLMs exhibit impressive language understanding and generation capabilities, their reasoning abilities—particularly for complex, multi-step problems—remain a critical challenge. In this talk, we explore a suite of advanced techniques to systematically improve LLM reasoning performance.
First, we discuss tool augmentation, where LLMs leverage external resources (e.g., Python interpreters) to improve the reasoning abilities. Next, we examine fine-tuning strategies, including supervised learning on interpolated reasoning datasets (e.g., via MATH or GSM8K). We highlight trade-offs between task-specific specialization and generalizability. We then present multi-agent frameworks, where multiple LLMs collaborate or debate to refine outputs. Techniques like self-consistency voting, iterative refinement, and role-specialized agents (e.g., "verifier" or "critic" models) mitigate individual model biases and errors. Our last focus is explicit reasoning traceability: structuring outputs as step-by-step chain-of-thoughts (CoT), where we introduce a learning from correctness way to trace the reasoning steps and correct possible wrong paths in the reasoning. Our results show promising solutions towards future LLMs with better reasoning abilities.
Bio: Prof. Linqi Song is an Associate Professor with the Department of Computer Science, City University of Hong Kong and a Research Scientist with the InnoHK AIFT Lab. He received the Ph.D. degree in electrical engineering from the University of California, Los Angeles (UCLA), USA and the B.S. and M.S. degrees from Tsinghua University, China. He was a Postdoctoral Scholar with the Department of Electrical and Computer Engineering, UCLA. His research interests include LLMs, reinforcement learning, federated learning, and information theory. He has published more than 200 journal and conference papers at main AI and IT venues, such as IEEE TIT, TPAMI, JSAC, TMC, JSTSP, NeurIPS, ICLR, and ACL. He has been the Associate Editor/Guest Editor for several journals, such as IEEE TIT, DSP, ICN, and he has been serving as TPC members for IEEE ISIT, ITW, Mobihoc, Infocom, etc. He has received the Best Paper Awards of IEEE MIPR 2020, the Best Paper Award of China Communications 2023, and three Silver Medals of Geneva International Exhibition of Inventions.