研究人员引入了Preisach Attention层（PAL），这是一种受数学物理中Preisach滞后算子启发的新的序列建模架构。PAL用二元继电器算子取代了标准的softmax attention，并将其状态维护为一个局部极值堆栈。这种新颖的方法在一个层中实现了图灵完备性，并能高效地计算历史范围统计数据，使其适用于需要具有弱位置依赖性的长时序记忆的任务。 AI
影响 引入了一种新颖的序列建模架构，有望提高需要长期记忆的任务的效率。
排序理由 该集群包含一篇详细介绍新AI架构的学术论文。
在 arXiv cs.NE (Neural & Evolutionary) 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2605.23603v1 Announce Type: cross Abstract: We introduce the Preisach Attention Layer (PAL), a novel sequence modelling architecture grounded in the classical Preisach hysteresis operator from mathematical physics. PAL replaces the softmax attention mechanism with a binary …
We introduce the Preisach Attention Layer (PAL), a novel sequence modelling architecture grounded in the classical Preisach hysteresis operator from mathematical physics. PAL replaces the softmax attention mechanism with a binary relay operator parameterised by learned activation…
We introduce the Preisach Attention Layer (PAL), a novel sequence modelling architecture grounded in the classical Preisach hysteresis operator from mathematical physics. PAL replaces the softmax attention mechanism with a binary relay operator parameterised by learned activation…