ConceptExtended Kronecker-Factored Approximate Curvature
EKFac
A method that approximates the Fisher information matrix of a neural network via Kronecker factorization. It is used to reduce the computational cost of second-order optimization and influence function calculations.
Timeline (last 90 days)
Oct 5
SSU-LSF uses dedicated EKFac influence functions for the Mamba state matrix to quantify the contribution of events to parameters via closed-form matrix exponential gradients.
Unverified50%