At the weekly Machine Learning Lunch Meetings, faculty members from Computer Science, Statistics, ECE, and other departments discuss their latest groundbreaking research in machine learning. This is an opportunity to network with faculty and fellow researchers, and to learn about the cutting-edge research being conducted in our university.
You can stay updated with the events hosted by MLLM by signing up to our mailing list at https://lists.cs.wisc.edu/mailman/listinfo/mllm.
Tuesdays 1:00-2:00pm in Computer Sciences 1221 unless noted otherwise. Co-organized by Jeremy McMahan
2/18 Grigorios Chrysos (ECE) Revisiting Character-level Adversarial Attacks for Language Models
2/25 Fred Sala (CS) Weak-to-Strong Generalization Through the Data-Centric Lens
3/4 Yiqiao Zhong (Statistics) Do You Interpret Your t-SNE Embeddings Correctly? A Perspective from Map-Continuity and Leave-One-Out
3/11 Qiaomin Xie (ISyE) Stochastic Methods in Variational Inequality Problems: Constant Stepsizes Go a Long Way
3/18 Manolis Vlatakis (CS) Solving Zero-Sum Convex Markov Games
4/1 Yudong Chen (CS) A Homotopy Optimization Approach to Jailbreak Attacks for LLMs
4/8 Ilias Diakonikolas (CS) Learning Multi-index Models
4/15 Kris Sankaran (Statistics) Enhancing Microbiome Analysis with Semisynthetic Data
4/22 Yongyi Guo (Statistics) Inference in Adaptive Experiments with Contextual Noise
Fridays 12:30-1:30pm in Computer Sciences 1221 unless noted otherwise. Co-organized by Jeremy McMahan
9/6 Jerry Zhu (CS) Game Changer: How to Minimally Modify a Two-Player Zero-Sum Game to Achieve Any Unique Nash Equilibrium
9/13 Sandeep Silwal (CS) Bi-metric framework for similarity search
9/20 Jordan Ellenberg (Math) What's up in applying machine learning to research mathematics?
9/27 Break --- No talk this week
10/4 Rob Nowak and Jifan Zhang (ECE) Show Me the Funny: LLM’s Epic Fail and the Road to Winning in the New Yorker Cartoon Caption Contest
10/11 Yiqiao Zhong (Statistics) Do LLMs solve novel tasks? An empirical investigation of out-of-distribution generalization
10/18 Grigorios Chrysos (ECE) Stairway to Specialization: The Path of Scalable Experts
10/25 (POSTPONED) Fred Sala (CS) Weak-to-Strong Generalization Through the Data-Centric Lens
11/1 Josiah Hanna (CS) Deploying reinforcement learning in the real-time domain of robot soccer
11/8 (ON ZOOM) Ramya Vinayak (ECE) PAL: Sample-Efficient Personalized Reward Modeling for Pluralistic Alignment
11/15 Kirthi Kandasamy (CS) Incentive-compatible Sequential Mechanisms with Strategic Non-myopic agents
11/22 Chaowei Xiao (iSchool) AI Security in the Era of Large Language Models and Agents
11/29 Break --- No talk this week
12/6 Yong Jae Lee (CS) How Intelligent Are Current Multimodal Video Models?
Thursdays 1-2pm in Computer Sciences 1221 unless noted otherwise. Co-organized by Jeremy McMahan
2/29 Grigoris Chrysos (ECE) Polynomial nets: Are activation functions required for learning in all deep networks?
3/7 Kirthi Kandasamy (CS) Data without Borders: Game-theoretic Challenges in Democratizing Data
3/14 cancelled due to speaker illness
3/21 (room: Researchers’ Link, WID 2nd floor) Kangwook Lee (ECE) Dual Operating Modes of In-Context Learning
4/4 Rob Nowak (ECE) What Kinds of Functions do Neural Networks Learn? Theory and Practical Applications
4/11 Kris Sankaran (Statistics) Transparent Synthetic Data Generation
4/18 Jessi Cisewski-Kehe (Statistics) Can ML help me find an exoplanet in my data?
4/25 (room: Biochemistry 1125) Yongyi Guo (Statistics) Balancing personalization and pooling: Decision-making and statistical inference with limited time horizons
5/2 Fred Sala (CS) Data and Compute-Efficient Foundation Model Adaptation
5/9 Junjie Hu (BMI) Aligning “Interlingual” Knowledge of Large Foundation Models
Thursdays 12-1pm in Computer Sciences 1325. Co-organized by Jeremy McMahan
12/7 Sharon Li (CS) Steering Large Language Models by Human Preferences
11/30 Qiaomin Xie (ISyE) Recent Advances in Average-Reward Restless Bandits
11/16 Kris Sankaran (STAT) Visualization in Deep Learning -- Theme and Variations
11/9 Dimitris Papailiopoulos (ECE) Can we teach addition to a small language model?
11/2 Yiqiao Zhong (STAT) A Geometric Journey into the world of large language models
10/26 Josiah Hanna (CS) On-policy Reinforcement Learning without On-policy Sampling
10/19 Yingyu Liang (CS) Provable Guarantees for Neural Networks via Gradient Feature Learning
10/12 Fred Sala (CS) How to Improve Your Zero-Shot Models, For Free
10/5 Yong Jae Lee (CS) Large Multimodal (Vision-Language) Models for Image Generation and Understanding
9/28 Young Wu + Jerry Zhu (CS) Two enemies are better than one
5/2 Junjie Hu (BMI) Towards A Better Understanding of Language Modeling and Reasoning
4/25 Mark Craven (BMI) Machine learning to Uncover Host-Virus Interactions
4/18 Rob Nowak (ECE) Active Machine Learning: Combining Human and Artificial Intelligence for Improved Learning Efficiency and Accuracy
4/11 Mohit Gupta (CS) Computer Vision, One Photon at a Time
3/28 Yingyu Liang (CS) Towards better understanding of deep learning: a perspective from data and algorithms
3/21 Yudong Chen (CS) Best of three worlds? Bias and extrapolation in constant-stepsize stochastic approximation
3/7 Dimitris Papailiopoulos (ECE) Looped Transformers as Programmable Computers
2/28 Josiah Hanna (CS) Scaling Off-Policy Evaluation to High-Dimensional State-Spaces Via State Abstraction
2/21 Kirthi Kandasamy (CS) Leveraging Reviews: Learning to Price with Buyer and Seller Uncertainty
2/14 Yong Jae Lee (CS) Large Multimodal (Vision-Language) Models for Image Generation and Understanding
2/7 Sharon Li (CS) (Data) Shift Happens, and How Should We Handle Them?
1/31 Jerry Zhu (CS) Inverse Nash Equilibrium
12/20 Fred Sala (CS) Weak Supervision for All Seasons
12/13 Ramya Vinayak (ECE) Learning from Diverse Data
11/22 Kangwook Lee (ECE) LIFT: Language-Interfaced FineTuning for Non-Language Machine Learning Tasks