Haocheng Xi(@HaochengXiUCB):Really exciting to see KV-cache compression getting attention. A similar bottleneck shows up beyond LLMs: for world models and autoregressive long-video generation, KV cache can quickly dominate memory and limit long-horizon consistency. Our recent work, Quant VideoGen, explores training-free 2-bit KV-cache quantization for video diffusion models, achieving up to 7.0× KV memory reduction with <4% latency overhead. Link:

Haocheng Xi

@HaochengXiUCB

Second-year PhD in @berkeley_ai. Prev: Yao Class, @Tsinghua_Uni | Efficient Machine Learning & ML sys

Joined August 2024

502 Following 2.3K Followers

Haocheng Xi@HaochengXiUCB

2026.03.27 03:29

Really exciting to see KV-cache compression getting attention. A similar bottleneck shows up beyond LLMs: for world models and autoregressive long-video generation, KV cache can quickly dominate memory and limit long-horizon consistency. Our recent work, Quant VideoGen, explores training-free 2-bit KV-cache quantization for video diffusion models, achieving up to 7.0× KV memory reduction with <4% latency overhead. Link: