注册并分享邀请链接,可获得视频播放与邀请奖励。

Oliver Sieberling @ COLM
@osieberling
PhD student @MIT, pretraining large neural networks, rnn enjoyer
加入 January 2024
794 正在关注    950 粉丝
Linear attention is arguably the most naive RNN, but still massively outperforms traditional RNNs by maintaining a matrix state. So... what if we use a (triadic) outer product of three vectors and maintain a three-dimensional state? Introducing: Triadic Linear Attention 🧵
显示更多
0
31
854
104
转发到社区