SPRING 2025
1) January 17th 2025 - Agenda Discussion
2) January 24th 2025 - KernMLOps (presented by Dr. Rossbach)
3) January 31st 2025 - Evolutionary Policy Optimization (presented by Lain)
4) February 7th 2025 - Cautious Optimizers (presented by Kaizhao)
5) February 14th 2025 - PrefixRL: Optimization of Parallel Prefix Circuits using Deep Reinforcement Learning (presented by Kayvan)
6) February 21st 2025 - Tensor Evolution (presented by Samarth)
7) February 28th 2025 - Monte Carlo Methods for Estimating Expectations and Gradients of Expectations in Reinforcement Learning (presented by Dr. Pingali)
8) March 7th 2025 Imitation Learning (presented by Lain)
9) March 7th 2025 - Reinforcement and Imitation Learning to Control Contact-Rich Interaction by Dr. Roberto Martín-Martín
10) March 21st 2025 - Spring Break - No Meeting
11) March 28th 2025 - No Meeting
12) April 4th 2025 - Physics-Informed Neural Networks (by Dr. Maciej Paszynski)
- Physics-informed machine learning (Karniadakis et al. 2021)
- Dr. Paszynski’s papers on PINNs:
- hp-VPINNs: Variational physics-informed neural networks with domain decomposition (Kharazmi et al. 2021)
- Slides
13) April 11th 2025 - Scheduling Languages: Past, Present & Future (presented by Dr. Elster)
14) April 18th 2025 - AI Coding Agents for Hardware-Optimized Code (presented by Dhairya, Irene, Pranoy)
- KernelBench: Can LLMs Write Efficient GPU Kernels?
- Baldur: Whole-Proof Generation and Repair with Large Language Models
- Slides
15) April 25th 2025 - The world’s simplest presentation of policy-gradient methods (presented by Dr. Pingali)
16) May 2nd 2025 - The world’s simplest presentation of baseline methods in policy-gradients (presented by Dr. Pingali)
17) May 16th 2025 - Measure once, optimize twice: trust regions and step-size restrictions in policy gradient methods (presented by William Ruys)