arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2109.05644cs.CL
收藏

SHAPE:用于Transformers的移位绝对位置嵌入

SHAPE: Shifted Absolute Position Embedding for Transformers

RIKEN(日本理化学研究所);Tohoku University(东北大学);Preferred Networks, Inc.(Preferred Networks公司)

Shun Kiyono, Sosuke Kobayashi, Jun Suzuki, Kentaro Inui

摘要:

位置表示对于在Transformers中构建位置感知表示至关重要。现有的位置表示遭受缺乏推广到具有不可见长度或高计算成本的测试数据。我们研究了移位绝对位置嵌入(SHAPE)来解决这两个问题。SHAPE的基本思想是通过在训练期间随机移动绝对位置来实现移位不变性,这是最近成功的位置表示的关键属性。我们证明,形状是经验可比的,它的对应,同时更简单,更快。

英文摘要:

Position representation is crucial for building position-aware representations in Transformers. Existing position representations suffer from a lack of generalization to test data with unseen lengths or high computational cost. We investigate shifted absolute position embedding (SHAPE) to address both issues. The basic idea of SHAPE is to achieve shift invariance, which is a key property of recent successful position representations, by randomly shifting absolute positions during training. We demonstrate that SHAPE is empirically comparable to its counterpart while being simpler and faster.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑