Multiway Multislice PHATE: Visualizing Hidden Dynamics of RNNs through Training

TMLR Paper5335 Authors

Abstract: 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 in understanding the functional principles of these networks and developing ideal model architectures and optimization strategies. Previous studies typically only emphasize the network representation post-training, overlooking their evolution process throughout training. Here, we present Multiway Multislice PHATE (MM-PHATE), a novel method for visualizing the evolution of RNNs' hidden states. MM-PHATE is a graph-based embedding using structured kernels across the multiple dimensions spanned by RNNs: time, training epoch, and units. We demonstrate on various datasets that MM-PHATE uniquely preserves hidden representation community structure among units and identifies information processing and compression phases during training. The embedding allows users to look under the hood of RNNs across training and provides an intuitive and comprehensive strategy for understanding the network's internal dynamics, such as why and how one model outperforms another or how specific architectures impact an RNN's learning ability.

Submission Length: Regular submission (no more than 12 pages of main content)

Assigned Action Editor: ~Russell_Tsuchida1

Submission Number: 5335