Attention Optimization - Dynamic sparse attention for accelerating long-context pre-filling.
Inference and Serving - Sparse attention calculation to reduce latency in long-context models.
[NeurIPS'24 Spotlight, ICLR'25, ICML'25] To speed up Long-context LLMs' inference, approximate and dynamic sparse calculate the attention, which reduces inference latency by up to 10x for pre-filling on an A100 while maintaining accuracy.
