多维多切片PHATE：通过训练可视化RNN的隐藏动态
Multiway Multislice PHATE: Visualizing Hidden Dynamics of RNNs through Training
AI总结：
MM-PHATE通过多维分析RNN隐藏状态演变，揭示训练过程中表示几何变化，提供直观方法理解RNN架构与性能关系。
摘要：
循环神经网络（RNN）是序列数据分析的常用工具，但常被视为黑箱。可视化RNN内部动态是理解其功能原理和开发更好架构的关键。本文提出MM-PHATE，一种基于图的嵌入方法，用于可视化RNN隐藏状态在时间、训练轮次和单元维度上的演变。在受控合成基准和真实RNN应用中，MM-PHATE保持单元间的隐藏表示社区结构，并揭示训练阶段的表示几何变化。在覆盖多种分岔家族和光滑状态空间扭曲的受控合成系统中，MM-PHATE恢复动态进展并区分家族级差异。在任务训练RNN中，嵌入识别训练过程中的信息处理和压缩相关阶段，时间分辨的几何和熵基总结与线性探测、时间步消融和标签-状态互信息一致。这些结果表明，MM-PHATE提供了一种直观且全面的方法，用于检查RNN隐藏动态并更好地理解模型架构和学习动态与性能的关系。
英文摘要：
Recurrent neural networks (RNNs) are a widely used tool for sequential data analysis; however, they are still often seen as black boxes. Visualizing the internal dynamics of RNNs is a critical step toward understanding their functional principles and developing better architectures and optimization strategies. Prior studies typically emphasize network representations only after training, overlooking how those representations evolve during learning. Here, we present Multiway Multislice PHATE (MM-PHATE), a graph-based embedding method for visualizing the evolution of RNN hidden states across the multiple dimensions spanned by RNNs: time, training epoch, and units. Across controlled synthetic benchmarks and real RNN applications, MM-PHATE preserves hidden-representation community structure among units and reveals training-phase changes in representation geometry. In controlled synthetic systems spanning multiple bifurcation families and smooth state-space warps, MM-PHATE recovers qualitative dynamical progression while distinguishing family-level differences. In task-trained RNNs, the embedding identifies information-processing and compression-related phases during training, and time-resolved geometric and entropy-based summaries align with linear probes, time-step ablations, and label--state mutual information. These results show that MM-PHATE provides an intuitive and comprehensive way to inspect RNN hidden dynamics across training and to better understand how model architecture and learning dynamics relate to performance.
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。
