Disentangle to Decay: Linear Attention with Trainable Decay Factor.
Haibo Tong, Chenyang Zhang, Jiayi Lin, Bingxuan Hou, Qingqing Hong, Junli Wang
Browse the full COLING paper archive.
Haibo Tong, Chenyang Zhang, Jiayi Lin, Bingxuan Hou, Qingqing Hong, Junli Wang
Browse the full COLING paper archive.