揭示大激活与注意力汇聚的机制差异及共现原因
The Spike, the Sparse and the Sink: Anatomy of Massive Activations and Attention Sinks
- 发现大激活是全局隐参数，注意力汇聚是局部依赖调制
- 预归一化结构导致两者共现，去除后二者解耦
- 适合研究Transformer内部机制的读者深入理解
我们研究了Transformer语言模型中两个常见现象：大激活（少数标记在少数通道中表现出极端异常值）和注意力汇聚（某些标记无论语义相关性如何都吸引大量注意力）。以往工作观察到二者常同时出现且涉及相同标记，但其功能角色和因果关系尚不明确。通过系统实验，我们发现这种共现主要源于现代Transformer架构的设计缺陷，且二者虽相关但功能不同：大激活具有全局效应，使隐藏表示在层间保持近似恒定，相当于模型的隐式参数；注意力汇聚具有局部效应，调节各注意力头的输出并偏向短程依赖。我们识别出预归一化配置是导致共现的关键因素，并证明移除该结构会使二者解耦。
原文摘要 · Abstract (English)
We study two recurring phenomena in Transformer language models: massive activations, in which a small number of tokens exhibit extreme outliers in a few channels, and attention sinks, in which certain tokens attract disproportionate attention mass regardless of semantic relevance. Prior work observes that these phenomena frequently co-occur and often involve the same tokens, but their functional roles and causal relationship remain unclear. Through systematic experiments, we show that the co-occurrence is largely an architectural artifact of modern Transformer design, and that the two phenomena serve related but distinct functions. Massive activations operate globally: they induce near-constant hidden representations that persist across layers, effectively functioning as implicit parameters of the model. Attention sinks operate locally: they modulate attention outputs across heads and bias individual heads toward short-range dependencies. We identify the pre-norm configuration as the key choice that enables the co-occurrence, and show that ablating it causes the two phenomena to decouple.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。