论文信息 - cosFormer: Rethinking Softmax in Attention - 字舞流文

cosFormer: Rethinking Softmax in Attention

Junjie Yan | Lingpeng Kong | Dongxu Li | Zhen Qin | Yiran Zhong | Weixuan Sun | Huicai Deng | Yunshen Wei | Baohong Lv