[Paper Reading] Viewing Log-Depth Transformers via the Lens of Distributed Computing
๐
Presented in OptiML Group Meeting (Fall 2024)
Posts
2024
[Paper Reading] StableSSM: Alleviating the Curse of Memory in State-space Models through Stable Reparameterization
๐
Presented in OptiML Group Meeting (Summer 2024)
[Paper Reading] Convex and Non-convex Optimization under Generalized Smoothness
๐
Presented in AI709 Advanced Deep Learning Theory (Spring 2024)
[Paper Reading] Understanding Gradient Descent on Edge of Stability in Deep Learning
๐
Presented in AI709 Advanced Deep Learning Theory (Spring 2024)
[Paper Reading] A U-turn on Double Descent: Rethinking Parameter Counting in Statistical Learning
๐
Presented in OptiML Group Meeting (Spring 2024)
2023
[Coursework Report] An Overview on Optimal Transport and its Application to Model Fusion
๐
Final report for the coursework โOptimal Transport + Gradient Flowโ
[Paper Reading] Implicit Bias of Large Depth Networks: a Notion of Rank for Nonlinear Functions
๐
Presented in OptiML Group Meeting (Fall 2023)
[Paper Reading] Adversarial training descends without descent: Finding actual descent directions based on Danskinโs Theorem
๐
Presented in OptiML Group Meeting (Spring 2023)
[Paper Reading] On Learning Fairness and Accuracy on Multiple Subgroups
๐
Presented in OptiML Group Meeting (Spring 2023)
2022
[Paper Reading] Offline Reinforcement Learning with Implicit Q Learning
๐
Presented in AI611 Deep Reinforcement Learning (Fall 2022)
[Paper Reading] WGAN with an In๏ฌnitely Wide Generator Has No Spurious Stationary Points
๐
Presented in OptiML Group Meeting (Spring 2022)
[Paper Reading] [KR only] Are Transformers universal approximators of sequence-to-sequence functions?
๐
๐ Summary: The first paper quantified Transformerโs expressive power