1 paper · 1 filter
Haoran Xu, Ziqian Liu, Rong Fu +5
With the evolution of large language models, traditional Transformer models become computationally demanding for lengthy sequences due to the quadratic growth in computation with r…