variational-methods

Conservative Lapse-Action Planning (CLAP): a variational access-and-dwell framework for safe AI optimization. Surf the conservative-lapse field to the best safe basin and dwell there — theorem-backed, pip-installable, with differentiable PyTorch training adapters.

Reference implementation of tensor-network variational diagonalization (TNVD): a spectrum MPS and finite-depth VQC jointly encode full many-body eigenenergies and eigenstates for ED-controlled validation and global spectral sampling beyond explicit enumeration.

Numerical image processing and geospatial raster workflows with TV reconstruction, sub-pixel registration, SAR primitives, GeoTIFF support, Docker, and reproducible benchmarks.