LGNov 3, 2025

The Curvature Rate λ: A Scalar Measure of Input-Space Sharpness in Neural Networks

arXiv:2511.01438v11 citations
Originality Highly original
AI Analysis

This provides a more interpretable and parameterization-invariant way to assess functional smoothness in neural networks, addressing a known bottleneck in sharpness metrics.

The paper tackled the problem of measuring curvature in neural networks by introducing a scalar curvature measure in input space, the curvature rate λ, which tracks high-frequency structure and can be shaped via regularization, achieving similar accuracy to Sharpness-Aware Minimization while improving geometry and calibration.

Curvature influences generalization, robustness, and how reliably neural networks respond to small input perturbations. Existing sharpness metrics are typically defined in parameter space (e.g., Hessian eigenvalues) and can be expensive, sensitive to reparameterization, and difficult to interpret in functional terms. We introduce a scalar curvature measure defined directly in input space: the curvature rate λ, given by the exponential growth rate of higher-order input derivatives. Empirically, λ is estimated as the slope of log ||D^n f|| versus n for small n. This growth-rate perspective unifies classical analytic quantities: for analytic functions, λ corresponds to the inverse radius of convergence, and for bandlimited signals, it reflects the spectral cutoff. The same principle extends to neural networks, where λ tracks the emergence of high-frequency structure in the decision boundary. Experiments on analytic functions and neural networks (Two Moons and MNIST) show that λ evolves predictably during training and can be directly shaped using a simple derivative-based regularizer, Curvature Rate Regularization (CRR). Compared to Sharpness-Aware Minimization (SAM), CRR achieves similar accuracy while yielding flatter input-space geometry and improved confidence calibration. By grounding curvature in differentiation dynamics, λ provides a compact, interpretable, and parameterization-invariant descriptor of functional smoothness in learned models.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes