March 2024 - Ai - Engineering Mailing Lists

Special AI Seminar: April 1, 2-3, Owen 102
by Tadepalli, Prasad 31 Mar '24

31 Mar '24

Hi all, We will have a special AI Seminar<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…> on Monday, April 1 at 2 pm on LLMs in Owen 102 by Bonnie Dorr. Please attend if you can. It will be counted towards AI507 credit. The Friday seminars will continue 2-3 starting April 5. Please register if you would like to get credits. Thank you, Prasad Tadepalli

1 1

Distinguished Lecture by Gesualdo Scutari (Purdue) 1pm, April 11
by Fu, Xiao 20 Mar '24

20 Mar '24

Dear all, Please mark your calendar for the distinguished lecture in the second week of the Spring Term! Cheers, Xiao Time: 1pm-2pm, April 11 Venue: KEC 1001 Title: Statistical Inference over Networks: Decentralized Optimization Meets High-dimensional Statistics Abstract: Abstract: There is growing interest in solving large-scale statistical learning problems over decentralized networks, where data are distributed across the nodes of the network and no centralized coordination is present (we termed these systems as “mesh” networks). Inference from massive datasets poses a fundamental challenge at the nexus of the computational and statistical sciences: ensuring the quality of statistical inference when computational resources, like time and communication, are constrained. While statistical-computation tradeoffs have been largely explored in the centralized setting, our understanding over mesh networks is limited: (i) distributed schemes, designed and performing well in the classical low-dimensional regime, can break down in the high-dimensional case; and (ii) existing convergence studies may fail to predict algorithmic behaviors, with some findings directly contradicted by empirical tests. This is mainly due to the fact that the majority of distributed algorithms have been designed and studied only from the optimization perspective, lacking the statistical dimension. This talk will discuss some vignettes from high-dimensional statistical inference suggesting new analyses (and designs) aiming at bringing statistical thinking in distributed optimization. Speaker: Bio: Gesualdo Scutari is a Professor with the School of Industrial Engineering and Electrical and Computer Engineering (by courtesy) at Purdue University, West Lafayette, IN, USA, and he is a Purdue Faculty Scholar. His research interests include continuous (distributed) optimization, equilibrium programming, and their applications to signal processing and statistical learning. Among others, he was a recipient of the 2013 NSF CAREER Award, the 2015 IEEE Signal Processing Society Young Author Best Paper Award, and the 2020 IEEE Signal Processing Society Best Paper Award. He serves as an IEEE Signal Processing Distinguish Lecturer (2023-2024). He served on the editorial broad of several IEEE journals and he is currently an Associate Editor of SIAM Journal on Optimization. He is a fellow of IEEE.

1 0

AI Seminar : Mar 15
by Mangannavar, Rajesh Devaraddi 13 Mar '24

13 Mar '24

Dear all, Our next AI seminar is scheduled to be on March 15th , 2-3 PM. It will be followed by a 30-minute Q&A session with the graduate students. Location: KEC 1001 Zoom link: https://oregonstate.zoom.us/j/96491555190?pwd=azJHSXZ0TFQwTFFJdkZCWFhnTW04U…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…> Hybrid RL: Using Offline Data for Efficient Online Reinforcement Learning Wen Sun Assistant Professor Computer Science Department Cornell University Abstract: Reinforcement Learning is a framework for learning to make sequential decisions from feedback. While most of today’s powerful AI systems (e.g., AlphaGo and ChatGPT) are trained via RL, RL is still considered to be inefficient in learning, i.e., it learns very slowly. In this talk, we consider hybrid RL, a new framework that integrates offline data into online RL with the goal of vastly improving the learning efficiency of RL. In the first part of the talk, we will present a simple approach to integrating offline data into a Q-learning-style framework. This simple approach achieves strong theoretical guarantees showing that with offline data, RL can be as easy as supervised learning. Empirically, when equipped with average-quality offline data, this approach can learn 10x faster than pure online RL baselines in a notorious hard video game that requires deep strategic exploration. In the second part of the talk, we show another way of integrating offline data into online RL via reset, a condition that allows a learning system to reset to any state it visited before and holds in important RL applications such as training language models. We propose an algorithm framework called dataset reset policy optimization (DrPO), which simply resets the RL agent to states in the offline data when collecting online data. We show that when integrating DrPO into an RL from Human Feedback (RLHF) framework where the reward model is learned from offline human preference data, it guarantees to discover the best policy covered by the offline human preference data. Empirically, when fine-tuning a 3 billion size language model on a standard RLHF benchmark — the Reddit TL;DR summarization task, DrPO learns to generate summaries that are better than that from humans, and a pure online RL baseline Proximal Policy Optimization (PPO) and a pure offline RL baseline Direct Preference Optimization (DPO), when evaluated by GPT4. Speaker Bio: Wen Sun is an Assistant Professor in the Computer Science Department at Cornell. Before that, he was a postdoctoral researcher at Microsoft Research NYC and he completed his Ph.D in 2019 from the Robotics Institute at Carnegie Mellon University. He is generally interested in machine learning, especially Reinforcement Learning. Much of his current research is about designing algorithms for efficient sequential decision-making and how to leverage human interaction for better and more efficient Reinforcement Learning. Please watch this space for future AI Seminars : https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…> Rajesh Mangannavar, Graduate Student Oregon State University ---- AI Seminar Important Reminders: -> For graduate students in the AI program, attendance is strongly encouraged

1 0

AI Seminar : Mar 08
by Mangannavar, Rajesh Devaraddi 06 Mar '24

06 Mar '24

Dear all, Our next AI seminar is scheduled to be on March 8th , 2-3 PM. It will be followed by a 30-minute Q&A session with the graduate students. Location: KEC 1001 Zoom link: https://oregonstate.zoom.us/j/96491555190?pwd=azJHSXZ0TFQwTFFJdkZCWFhnTW04U…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…> >From Large to Small Datasets: Size Generalization for Clustering Algorithm Selection Ellen Vitercik Assistant Professor Management Science & Engineering & Computer Science Stanford University Abstract: In clustering algorithm selection, we are given a massive dataset and must efficiently select which clustering algorithm to use. We study this problem in a semi-supervised setting, with an unknown ground-truth clustering that we can only access through expensive oracle queries. Ideally, the clustering algorithm's output will be structurally close to the ground truth. We approach this problem by introducing a notion of size generalization for clustering algorithm accuracy. We identify conditions under which we can (1) subsample the massive clustering instance, (2) evaluate a set of candidate algorithms on the smaller instance, and (3) guarantee that the algorithm with the best accuracy on the small instance will have the best accuracy on the original big instance. We verify these findings both theoretically and empirically. Speaker Bio: Ellen Vitercik is an Assistant Professor at Stanford University with a joint appointment between the Management Science & Engineering department and the Computer Science department. Her research revolves around machine learning theory, discrete optimization, and the interface between economics and computation. Before joining Stanford, she spent a year as a Miller Fellow at UC Berkeley after receiving a PhD in Computer Science from Carnegie Mellon University. Her thesis won the SIGecom Doctoral Dissertation Award and the CMU School of Computer Science Distinguished Dissertation Award. Please watch this space for future AI Seminars : https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…> Rajesh Mangannavar, Graduate Student Oregon State University ---- AI Seminar Important Reminders: -> For graduate students in the AI program, attendance is strongly encouraged

1 0

AI Seminar: March 1
by Tadepalli, Prasad 01 Mar '24

01 Mar '24

Dear all, Tomorrow's AI seminar will be by Yoon Kim, who will be talking about large language models. This seminar will be remote, but please attend in person if possible. Prasad Location: KEC 1001 Zoom link: https://oregonstate.zoom.us/j/96491555190?pwd=azJHSXZ0TFQwTFFJdkZCWFhnTW04U…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…> Large Language Models and Symbolic Structures Yoon Kim EECS, Massachusetts Institute of Technology Abstract: Over the past decade the field of NLP has shifted from a pipelined approach (wherein intermediate symbolic structures such as parse trees are explicitly predicted and utilized for downstream tasks) to an end-to-end approach wherein pretrained large language models (LLMs) are adapted to various downstream tasks via finetuning or prompting. What role (if any) can symbolic structures play in the era of LLMs? In the first part of the talk, we will see how latent symbolic structures in the form of hierarchical alignments can be used to guide LM-based neural machine translation systems to improve translation of low resource languages and even enable the use of new translation rules during inference. In the second part, we will see how expert-derived grammars can be used to control LLMs via prompting for tasks such as semantic parsing where the output structure must obey strict domain-specific constraints. Speaker Biography: Yoon Kim is an assistant professor at MIT EECS. He received his PhD from Harvard University, advised by Alexander Rush.

3 3