RT-Transformer：Transformer块作为球面状态估计器
The Transformer as a Polar State Estimator
AI总结：
本文提出RT-Transformer，通过几何估计问题自然推导出Transformer核心组件，实现状态估计的球面建模与高效更新。
摘要：
本文展示Transformer块的核心组件——注意力、残差连接和归一化——源于单一的几何估计问题。将潜在状态建模为超球面上的方向，噪声定义在当前估计点的切平面中，从而得到一种精度加权的方向推断过程，其中注意力聚合证据，残差连接实现增量状态更新，归一化将更新后的状态重新投影到超球面上。这些组件源于估计问题的几何特性，而非作为独立的架构选择引入。
英文摘要：
We show that the core components of the Transformer---attention, residual connections, and normalization---arise naturally from a single geometric state estimation problem. Modeling the latent state in polar coordinates naturally separates radial and hyperspherical dynamics, yielding a precision-weighted filtering procedure in which normalization enforces the hyperspherical constraint, attention aggregates directional evidence, and the residual connection implements an incremental state update. The standard Transformer block with rotary positional encodings is recovered by discarding the geometric correction terms of the resulting state estimator, showing that its architecture follows from the underlying estimation problem rather than from independent design choices. The proposed \textit{Polar Transformer} retains these geometric corrections.
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。
