5 papers · 1 filter
Memory as Weights: Internalizing Long-Term History for Streaming Videos
Zhuoming Liu, Jinhong Lin, Lin Zhang +2
Streaming vision-language models (VLMs) must continuously integrate information from the past while remaining responsive to current observations and user queries. Existing memory m…
Pulse: Training Acceleration for Large Diffusion Models with Automatic Pipeline Parallelism
Boran Sun, Guoyong Jiang, Lin Zhang +8
Diffusion models are now a dominant approach for high-fidelity image and video generation, yet scaling their training across GPU clusters remains challenging. Unlike transformer-on…
UniTemp: Unlocking Video Generation in Any Temporal Order via Bidirectional Distillation
Lin Zhang, Sicheng Mo, Zefan Cai +6
Autoregressive video diffusion models have emerged as a promising approach for long video generation, achieving strong performance in streaming settings. However, existing methods…
DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models
Zhuoming Liu, Jinhong Lin, Kwan Man Cheng +3
Many modern vision-language models (VLMs) build on autoregressive decoding of discrete tokens. While text-based output interfaces enable scalable pretraining and strong zero-shot g…
Data Warmup: Complexity-Aware Curricula for Efficient Diffusion Training
Jinhong Lin, Pan Wang, Zitong Zhan +2
A key inefficiency in diffusion training occurs when a randomly initialized network, lacking visual priors, encounters gradients from the full complexity spectrum--most of which it…