Skip to content

Temporal Functional Circuits

Home / Concepts / Temporal Functional Circuits

Temporal Functional Circuits

Auto-generated stub. Edit this file to add more details.

A framework for interpreting Kolmogorov-Arnold Network edge functions in time-series forecasting through temporal attribution and edge-level intervention analysis.

Why It Matters

It enables mechanistic interpretability in KAN-based time-series forecasting by mapping edge functions to specific input lags and verifying their predictive contribution.

Evidence

This paper introduces Temporal Functional Circuits, a framework that transforms KAN edge functions from latent visualizations into faithful, temporally grounded explanations.

Metadata & Links