揭示注意力机制中归一化导致信息区分能力下降的问题
Limitations of Normalization in Attention Mechanism
- 从几何角度分析注意力选择机制，给出向量分离的理论边界
- 实验证明选更多词时模型区分力下降，趋向均匀选择
- 适合关注注意力机制原理和改进方向的研究者
本文研究注意力机制中归一化的局限性。基于理论框架，我们识别出模型的选择能力及标记向量间的几何分离特性，给出了在softmax缩放下向量距离与分离标准的显式界限。通过预训练GPT-2模型的实验，我们验证了理论结果并分析了注意力机制的关键行为。结果显示，随着被选标记数量增加，模型区分有效标记的能力显著下降，常趋于均匀选择模式。此外，我们还发现软最大归一化下的梯度敏感性在低温度设置下对训练构成挑战。这些发现深化了对基于softmax注意力机制的理解，并推动未来注意力架构采用更稳健的归一化与选择策略。
原文摘要 · Abstract (English)
This paper investigates the limitations of the normalization in attention mechanisms. We begin with a theoretical framework that enables the identification of the model's selective ability and the geometric separation involved in token selection. Our analysis includes explicit bounds on distances and separation criteria for token vectors under softmax scaling. Through experiments with pre-trained GPT-2 model, we empirically validate our theoretical results and analyze key behaviors of the attention mechanism. Notably, we demonstrate that as the number of selected tokens increases, the model's ability to distinguish informative tokens declines, often converging toward a uniform selection pattern. We also show that gradient sensitivity under softmax normalization presents challenges during training, especially at low temperature settings. These findings advance current understanding of softmax-based attention mechanism and motivate the need for more robust normalization and selection strategies in future attention architectures.
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。