ridge_mkl_superblock — n4m.compose.aom_superblock.ridge_mkl_superblock

Namespace: n4m.compose.aom_superblock · Fully-qualified: n4m.compose.aom_superblock.ridge_mkl_superblock · Catalog id: aom_pop.ridge_mkl_superblock

C ABI symbols (ABI 2.0): n · o · n · e

Python: from n4m.compose.aom_superblock import aom_ridge_mkl_superblock

Python-backed donor-style AOM Ridge MKL-light superblock constrained to strict-linear single-operator AOM views. It learns non-negative train-only KTA weights over operator blocks inside every alpha-CV fold, refits weights on the full calibration set, fits native Ridge on the equivalent weighted superblock, and folds final coefficients back to original-input input_coefficients plus intercept. It intentionally excludes donor branch_global, row-reference-dependent preprocessing, nonlinear kernels and nonlinear AOM Ridge modes; native v1 builds in CUDA-enabled configurations but this is not yet a fused GPU weighted-superblock grinder.

Timing benchmark: benchmarks/cross_binding/bench_aom_ridge_mkl_superblock_timing.py

See also: methods index.