2 papers
cs.LG2026
When Denoising Hurts: Rethinking the Terminal Step of Diffusion Time Series Forecasters -- Extended Version
Dat Nguyen-Cong, Luong Tran, Tung Kieu
Diffusion models offer a natural way to model uncertainty in time series forecasting, yet their iterative sampling process is often treated as a uniformly beneficial refinement pro…
cs.CL2026
FastDiSS: Few-step Match Many-step Diffusion Language Model on Sequence-to-Sequence Generation--Full Version
Dat Nguyen-Cong, Tung Kieu, Hoang Thanh-Tung
Self-conditioning has been central to the success of continuous diffusion language models, as it allows models to correct previous errors. Yet its ability degrades precisely in the…