1 paper · 1 filter
Xin Wang, Hui Shen, Boyuan Zheng +7
The scalability of Large Language Models (LLMs) to long contexts is fundamentally constrained by the quadratic complexity of standard attention, motivating the adoption of linear a…