7 papers
Principled Top- Selection for Language Models with Hybrid Gradients
Xuchen Gong, Junfei Sun, Tian Li
Selecting the best items out of candidates is a critical component of modern large language model systems, such as document selection in Retrieval-Augmented Generation (RAG…
Asynchronous Heavy-Tailed Optimization
Junfei Sun, Dixi Yao, Xuchen Gong +3
Heavy-tailed stochastic gradient noise, commonly observed in transformer models, can destabilize the optimization process. Recent works mainly focus on developing and understanding…
Saving Foundation Flow-Matching Priors for Inverse Problems
Yuxiang Wan, Ryan Devera, Wenjie Zhang +1
Foundation flow-matching (FM) models promise universal priors for solving inverse problems (IPs); yet today, they trail behind domain-specific and even untrained priors. \emph{How…
FMPlug: Plug-In Foundation Flow-Matching Priors for Inverse Problems
Yuxiang Wan, Ryan Devera, Wenjie Zhang +1
We present FMPlug, a novel plug-in framework that enhances foundation flow-matching (FM) priors for solving ill-posed inverse problems. Unlike traditional approaches that rely on d…
Temporal-Consistent Video Restoration with Pre-trained Diffusion Models
Hengkang Wang, Yang Liu, Huidong Liu +5
Video restoration (VR) aims to recover high-quality videos from degraded ones. Although recent zero-shot VR methods using pre-trained diffusion models (DMs) show good promise, they…
DMPlug: A Plug-in Method for Solving Inverse Problems with Diffusion Models
Hengkang Wang, Xu Zhang, Taihui Li +3
Pretrained diffusion models (DMs) have recently been popularly used in solving inverse problems (IPs). The existing methods mostly interleave iterative steps in the reverse diffusi…
What is Wrong with End-to-End Learning for Phase Retrieval?
Wenjie Zhang, Yuxiang Wan, Zhong Zhuang +1
For nonlinear inverse problems that are prevalent in imaging science, symmetries in the forward model are common. When data-driven deep learning approaches are used to solve such p…