LayerRoPE：将 Transformer 深度方向的位置编码显式化

1. 12:00
   arXiv · Machine Learning Theory

   LayerRoPE：动态深度幅度与角度叠加

   研究者提出 LayerRoPE，将 Transformer 隐藏状态范数随深度增长的现象解释为涌现的深度位置编码，并用单一共享向量与深度条件标量替换逐层归一化权重 $\gamma$，在参数净减少、FLOPs 变化 $<0.02\%$ 的条件下实现深度轴上的隐式 RoPE。

Current heat 9·Comparable peak 10（Oct 8）·Comparable change over 24 hours –

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.