Disen Lan

MoM: Linear Sequence Modeling with Mixture-of-Memories

Go with Your Gut: Scaling Confidence for Autoregressive Image Generation

Improving Bilinear RNN with Closed-loop Control

Liger: Linearizing Large Language Models to Gated Recurrent Structures

TimeFilter: Patch-Specific Spatial-Temporal Graph Filtration for Time Series Forecasting

LASP-2: Rethinking Sequence Parallelism for Linear Attention and its Hybrid