TILOS Seminar Series

Events

Events Search and Views Navigation

Event Views Navigation

Today

  • Headshot of Dr. Volkan Cevher

    Optimization for AI and ML Seminar: Training Neural Networks at Any Scale

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Volkan Cevher, École Polytechnique Fédérale de Lausanne Abstract: At the heart of deep learning’s transformative impact lies the concept of scale--encompassing both data and computational resources, as well as their interaction with neural network architectures. Scale, however, presents critical challenges, such as increased instability during training and prohibitively expensive model-specific tuning. Given the substantial resources […]

  • Headshot of Dr. Frank Curtis

    Optimization for ML and AI Seminar: Stochastic-Gradient and Diagonal-Scaling Algorithms for Constrained Optimization and Learning

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Frank E. Curtis, Lehigh University Abstract: I will motivate and provide an overview of recent efforts in my research group on the design and analysis of stochastic-gradient-based algorithms for solving constrained optimization problems. I will focus in particular on our motivation for informed supervised learning, where constraints in the training problem can be used to […]

  • Headshot of Dr. Yi Wu

    TILOS-HDSI Seminar: Incentivizing Emergent Behaviors for LLMs via Reinforcement Learning

    Qualcomm Conference Center Room B (Jacobs Hall first floor) and Virtual 9736 Engineers Ln, La Jolla, CA, United States

    Yi Wu, Tsinghua University Abstract: Reinforcement Learning (RL) has become a powerful post-training method for eliciting advanced behaviors in large language models (LLMs). This talk presents recent results showing how RL can incentivize the emergence of LLM capabilities across three domains: (1) multi-player deduction game, Werewolf, where RL-trained LLM agents develop strategic behaviors and outperform […]

  • Headshot of Dr. Joel Tropp

    Optimization for ML and AI Seminar: Randomized linear algebra with subspace injections

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Joel Tropp, Caltech Abstract: To achieve the greatest possible speed, practitioners regularly implement randomized algorithms for low-rank approximation and least-squares regression with structured dimension reduction maps. This talk outlines a new perspective on structured dimension reduction, based on the injectivity properties of the dimension reduction map. This approach provides sharper bounds for sparse dimension reduction […]

  • Headshot of Dr. Koushil Sreenath

    TILOS-HDSI Seminar: Safety, Representations, and Generative Learning in Dynamical Systems

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Koushil Sreenath, UC Berkeley Abstract: This talk explores the interplay between model-based guarantees and learning-based flexibility in the control of dynamical systems. I begin with safety-critical control using control barrier functions (CBFs), highlighting that while CBFs enforce state constraints, they may induce unstable internal dynamics. I introduce conditions under which CBF-based safety filters ensure boundedness […]

  • Headshot of Dr. Tengyu Ma

    Canceled [CANCELED] Optimization for ML and AI Seminar: Fantastic Pretraining Optimizers and Where to Find Them

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Tengyu Ma, Stanford Abstract: AdamW has long been the dominant optimizer in language model pretraining, despite numerous claims that alternative optimizers offer 1.4 to 2x speedup. We posit that two methodological shortcomings have obscured fair comparisons and hindered practical adoption: (i) unequal hyperparameter tuning and (ii) limited or misleading evaluation setups. To address these two […]

  • Headshot of Dr. Yang Zheng

    Optimization for ML and AI Seminar: Extended Convex Lifting for Policy Optimization in Control

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Yang Zheng, UC San Diego Abstract: Direct policy search has achieved great empirical success in reinforcement learning. Many recent studies have revisited its theoretical foundation for continuous control, which reveals elegant nonconvex geometry in various benchmark problems. In this talk, we introduce an Extended Convex Lifting (ECL) framework, which reveals hidden convexity in classical optimal […]

  • Headshot of Dr. Sung-Kyu Lim

    TILOS-MICS Seminar: AI-Driven Design Automation for Multi-Chip Integration in AI Chips

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Sung-Kyu Lim, University of Southern California Abstract: Multi-chip integration has become a standard approach in AI training and is rapidly gaining traction in edge learning applications. Leveraging 2.5D and 3D IC architecture enables substantial improvements in energy efficiency and latency by optimizing inter chip data transfer. At the core of this transformation lies the automation […]

  • Headshot of Dr. Maarten de Hoop

    TILOS-HDSI Seminar: Kinetic Theory Perspective of Foundation Models for Physics

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Maarten de Hoop, Rice University Abstract: We present a kinetic theory perspective of foundation models for physics. We begin with providing a mathematical framework for analyzing transformers. To uniformly address their expressivity, we consider the case that the mappings are conditioned on a context represented by a probability distribution of tokens. That is, transformers become […]

  • Headshot of Dr. Jason Eshraghian

    TILOS-HDSI Seminar: Neuromorphic LLMs

    HDSI 123 and Virtual 3234 Matthews Ln, La Jolla, CA, United States

    Jason Eshraghian, UC Santa Cruz Abstract: This talk will show you what neuromorphic computing can do when an academic lab accidentally pulls $2-million of GPU-hours. We will showcase a series of frontier reasoning LLMs developed out of an academic lab, from data curation and pre-training to post-training and alignment. These models surpass leading LLMs from […]