Algorithm Development in Neural Networks: Insights from the Streaming Parity Task

Loek van Rossem, Andrew Saxe

International Conference on Machine Learning 2025 · Oral

In this insightful talk from ICML 2025, Loek van Rossem and Andrew Saxe delve into the fascinating phenomenon of how neural networks implicitly learn computational algorithms from training data, leading to remarkable generalization capabilities. The presentation focuses on a specific, yet profoundly illustrative, problem: the **streaming parity task**. The core challenge addressed is understanding the bridge between the continuous dynamics of gradient descent optimization and the emergent, discrete computational algorithms that enable neural networks to solve tasks far beyond the scope of their explicit training.

AI review

Van Rossem and Saxe present a mechanistic account of how RNNs trained on the streaming parity task transition from sequence-fitting to finite-state computation via representation mergers. The central contribution — a low-dimensional ODE system derived from a linearized abstraction of the optimization dynamics, whose fixed point predicts whether two internal representations will coalesce — is a genuine theoretical step toward explaining a phenomenon the community has been observing empirically. The work is honest about its scope, which is narrow (regular languages, RNNs, one synthetic task)…