..................................................................................3/19/2025..................................................................................
Zhihan Zhang is a fourth-year Ph.D. student at the University of Notre Dame focusing on instruction tuning and large language models (LLMs), with 6 years of experience in the field of natural language processing (NLP). Before his Ph.D., Zhihan received his B.S. from Peking University. Zhihan has published over 30 papers, including nine as the first author, in top-tier machine learning and NLP venues such as ACL, EMNLP, NAACL, ICLR, and TACL. He is also leading a tutorial on instruction tuning at EMNLP 2025.
Title: Instruction Tuning: The Road to Intelligent LLMs
Abstract: Large language models (LLMs) have demonstrated strong performance across numerous NLP tasks, yet they struggle to interact seamlessly with human users. Instruction tuning addresses this gap by training LLMs to understand and follow natural language instructions, which helps them evolve into intelligent AI assistants. In this lecture, I will first introduce the technical motivations behind instruction tuning. Next, I will explore how the instruction-following ability is evaluated and highlight some state-of-the-art techniques used in instruction tuning LLMs. Finally, I will discuss a few promising research directions aimed at developing next-generation instruction-following models.
..................................................................................2/21/2025..................................................................................
Dr. Lili Mou is an Assistant Professor at the Department of Computing Science, University of Alberta. He is also an Alberta Machine Intelligence Institute (Amii) Fellow and a Canada CIFAR AI (CCAI) Chair. Lili received his BS and PhD degrees in 2012 and 2017, respectively, from the School of EECS, Peking University. After that, he worked as a postdoctoral fellow at the University of Waterloo. His research interests mainly lie in designing novel machine learning algorithms and frameworks for NLP. He has publications at top conferences and journals, including ACL, EMNLP, TACL, ICML, ICLR, and NeurIPS. He also presented tutorials at EMNLP'19 and ACL'20. He received an AAAI New Faculty Highlight Award in 2021.
Title: Multi-Teacher Distillation: An Ensemble-Then-Distill Approach
Abstract: Knowledge distillation (KD) aims to transfer the knowledge in a large model (called a teacher) into a small one (called a student), and has become an emerging research topic as the sizes of deep learning models keep growing. Today, there are abundant readily available large models, such as ChatGPT, LLaMa, and T5. It then becomes natural to ask: Can we distill the knowledge from multiple teachers? At first glance, it appears easy to perform multi-teacher KD, as we can simply train the student from the union of teachers’ predictions. However, I would argue that such a naïve attempt may not work well for multi-teacher KD. This is because traditional KD adopts the cross-entropy loss, which tends to yield a smooth distribution. In this talk, I will present a novel ensemble-then-distill approach, which builds an ensemble of teacher models to train the student. I will also discuss applications to text generation and syntactic parsing.
..................................................................................1/24/2025..................................................................................
Fei Wang is a Ph.D. candidate in computer science at the University of Southern California. His research focuses on identifying and mitigating fundamental risks in the development and deployment of generative AI models to enhance their robustness, controllability, and safety, ensuring trustworthy and responsible outcomes. His work has been recognized with an Amazon ML PhD Fellowship, an Annenberg PhD Fellowship, a USC Graduate School Research Award, and a USC CS Best Research Award. Additionally, he co-instructed a recent tutorial at EMNLP 2024 titled Enhancing LLMs’ Capabilities Beyond Scaling-up.
Title: From Risk to Resilience: Navigating Misalignment in Large Language Models
Abstract: As large language models (LLMs) become central to intelligent systems, their use is expanding from everyday applications to high-stakes domains. Alignment plays a crucial role in the success of LLMs, ensuring that model behavior matches our expectations and remains consistent with various objectives. However, misalignment remains a significant challenge that undermines the trustworthiness and responsibility of these models. This talk will explore methods to navigate the misalignment problem by addressing three key research questions: (1) How to mitigate the risk of a misaligned LLM when only limited inference access is available? (2) How to ensure a reliable alignment process in multimodal scenarios? (3) How to integrate missing or customized alignment objectives to achieve precise control over model behavior in diverse contexts.
..................................................................................11/21/2024..................................................................................
Benjamin Feuer is a Ph.D. candidate in the Department of Computer Science and Engineering at NYU. He is a member of the DICE Lab and an active collaborator with AI startups Arthur.AI and Abacus.AI. Previously, he received his MS in Computer Science from New York University (2022); prior to that, he worked as a professional screenwriter and film director. His recent research interests include data-centric factors in machine learning systems, robust LLM benchmarking, evaluation and alignment, and scalable data integration for very large databases.
Title: Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarking
Abstract: The release of ChatGPT in November 2022 sparked an explosion of interest in LLM alignment with human values, preferences and standards. Existing methods claim superiority by virtue of better correspondence with human pairwise preferences, often measured by LLM judges. But do LLM-judge preferences translate to progress on other, more concrete metrics for alignment, and if not, why not? Recent joint research with NYU, Columbia and Arthur.AI shows that (1) LLM-judgments do not correlate with concrete measures of safety, world knowledge, and instruction following; (2) LLM judges have powerful implicit biases, prioritizing style over factuality and safety; and (3) the supervised fine-tuning (SFT) stage of post-training, rather than RLHF, has the greatest impact on objective measures of alignment.
..................................................................................9/24/2024..................................................................................
Wu Lin is a postdoctoral fellow at the Vector Institute, which Geoffrey Hinton founded. He got a Ph.D. degree from The University of British Columbia in 2023. His research direction is practical geometric methods for large-scale optimization. He has published more than ten papers on optimization at top machine-learning conferences. Currently, he is working on training deep neural networks such as transformers.
Title: Connections Between Approximate Inference and Adaptive Optimization
Abstract: He will talk about designing training methods for large neural networks.
..................................................................................4/26/2024..................................................................................
Qingyun Wang is a CS Ph.D. student at UIUC, advised by Heng Ji. His research lies in controllable knowledge-driven natural language generation, focusing on NLP for scientific discovery. He served as a PC member for multiple conferences including ICML, ACL, ICLR, NeurIPS, etc. He previously entered the finalist of the first Alexa Prize competition. He received the NAACL-HLT 2021 Best Demo Award. Qingyun will join the Data Science Department at William & Mary as an Assistant Professor, starting in August 2025.
Title: AIScientist: Toward Automated Literature Understanding and Scientific Discovery
Abstract: Due to the rapid growth of publications varying in quality, there exists a pressing need to help scientists digest and evaluate relevant papers, thereby facilitating scientific discovery. This talk aims to provide an overview of the AI-assisted scientific paper lifecycle, detailing how machines can augment every stage of the research process for the scientist, including scientific literature understanding, experiment development, manuscript draft writing, and finally draft evaluation. I will first cover fine-grained few-shot scientific entity extraction and then show how the extracted knowledge graph can be applied to scientific literature review and scientific hypothesis discovery.