sparse-attention 12 large-language-models 3 attention 3 low-rank 2 research 2 vision-transformer 2 attention-mechanism 2 natural-language-processing 2 deep-learning 2 long-context 1 artificial-intelligence 1 model-compression 1 lightweight 1 super-resolution 1 text-summarization 1 efficient-transformers 1 efficient-vision-transformers 1 fast-inference 1 sparse-neural-networks 1 llm-inference 1 high-throughput 1 cpu-offload 1 vit 1 video-generation 1 sageattention 1 quantization 1 mlsys 1 llm 1 inference-acceleration 1 ai-infra 1 inference-optimization 1 efficient-ai 1 pytorch 1 neighborhood-attention 1 cuda 1