Neural Attention Search Linear: Towards Adaptive Token-Level Hybrid Attention Models 事件

Neural Attention Search Linear: Towards Adaptive Token-Level Hybrid Attention Models arXiv:2602.03681v2 Announce Type: replace Abstract: The quadratic computational complexity of softmax transformers has become a bottleneck in long-context scenarios. In contrast, linear attention model families provide a promising direction towards a more efficient sequential model. These linear attention models compress past KV values into a single hidden state, thereby efficiently reducing complexity during b