pith. sign in
module module high

IndisputableMonolith.Cost.Ndim.Metric

show as:
view Lean formalization →

This module supplies the metric tensor entries obtained by taking the Hessian of the n-dimensional reciprocal cost function JlogN expressed in logarithmic coordinates. Researchers modeling the geometry of cost landscapes in Recognition Science would cite these definitions when assembling the full metric from second derivatives. The construction reduces the multidimensional Hessian to a form depending on a single weighted dot product, as prepared in the imported Hessian module.

claimThe metric entry derived from the Hessian of the cost $JlogN$ in log coordinates is $m_{ij} = Hess(JlogN)_{ij}$, where the n-dimensional cost depends only on the aggregate $dot α t$.

background

The module sits inside the Cost.Ndim hierarchy and imports the Hessian formulas for the n-dimensional reciprocal cost. In log-coordinates the n-dimensional cost depends only on the single weighted aggregate dot α t, so its Hessian reduces to a rank-1 update controlled by that scalar. Sibling definitions metricEntry, metricEntry_zero and metric_at_equilibrium_eq_hessian then package the resulting second-derivative entries for direct use in the metric tensor.

proof idea

This is a definition module, no proofs.

why it matters in Recognition Science

The module completes the concrete metric construction for JlogN inside the n-dimensional cost framework, supplying the Hessian-derived entries that higher-level geometry and equilibrium statements rely on. It directly implements the reduction to the aggregate dot α t that the upstream Hessian module isolates, thereby closing the interface between the abstract cost function and its Riemannian structure.

scope and limits

depends on (1)

Lean names referenced from this declaration's body.

declarations in this module (3)