3 papers
cs.LG2026
Spectral Asymptotics of Neural Network Loss Landscapes: An Exact Decomposition of the Curvature Exponent
Anherutowa Calvo
The curvature exponent in -- governing how Hessian eigenvalues scale with gradient singular values -- varies systematically across layer types ($α\appro…
cs.LG2026
Predictable Gradient Manifolds in Deep Learning: Temporal Path-Length and Intrinsic Rank as a Complexity Regime
Anherutowa Calvo
Deep learning optimization exhibits structure that is not captured by worst-case gradient bounds. Empirically, gradients along training trajectories are often temporally predictabl…
cs.LO2026
Chronology as a Consistency Invariant in Composable Information Systems
Anherutowa Calvo, Dante K. Calvo
We formalize a minimal setting in which a chronology (a strict partial order on events) is forced by consistency of distributed information under local composability. The system ma…