SFT Torch MIT-Style Attribution and Research IP License

Copyright (c) 2026 Dmitry Sierikov

Project and attribution link: https://github.com/dimaq12/spectral_flow

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to use,
copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
the Software, and to permit persons to whom the Software is furnished to do
so, for private, academic, research, commercial, and proprietary purposes,
subject to the following conditions:

1. The above copyright notice, this license text, and the project attribution
   link must be included in all copies or substantial portions of the Software.

2. If the Software or a substantial portion of it is redistributed in source
   form, binary form, package form, documentation, model/tooling form, or a
   public product that includes third-party notices, the following attribution
   must be reasonably visible in the relevant license, notice, documentation,
   credits, or about material:

   Spectral Flow Transform / SFT Torch
   Copyright (c) 2026 Dmitry Sierikov
   https://github.com/dimaq12/spectral_flow

3. The name of Dmitry Sierikov may not be used to imply endorsement of any
   derived work, product, service, or organization without prior written
   permission.

Research IP and Theory Attribution Notice
-----------------------------------------

The Software contains original research ideas, mathematical architecture,
experimental theory, and implementation designs authored by Dmitry Sierikov.
This license grants permission to use the Software as stated above, but it
does not waive attribution to Dmitry Sierikov for the original research
contributions embodied in this repository.

The protected attribution scope includes the novel synthesis, terminology,
software architecture, experiments, validation gates, and theory-to-code
realizations expressed in this repository, including but not limited to:

- The Spectral Flow Transform (SFT) software architecture for data-to-operator
  compilation and W-centric spectral programming.
- The use of Hellmann-Feynman kernels as a universal local spectral interface
  for prediction, inverse design, codec-style transforms, and data-agnostic
  operator programs.
- The layered DataView / Lift / OperatorSpec / ExecutionPlan / SpectralProgram
  API design for compiling heterogeneous signals, graphs, grids, sequences,
  tokens, and explicit operators into a common spectral operator workflow.
- The integration of Omega and S-operator tail diagnostics with SFT spectral
  programs, including composition-law oriented validation gates and
  finite-N convergence diagnostics.
- The noncommutative Hermitian continuation layer, including commutator
  diagnostics and segmented W-refresh prediction controls.
- The non-self-adjoint SFT stack, including biorthogonal W, biorthogonal
  Hessian validation, cheap C_H prime proxy diagnostics, conditioned F-prime
  continuation, and proxy-based step/refresh control as realized here.
- The experimental exceptional-point workflow, including EP atlases,
  projector/distance branch tracking, Puiseux-coordinate inverse helpers,
  Jordan fusion/fingerprints, and negative-control validation gates.
- The performance guardrail design for keeping expensive exact Hessian and
  pseudospectrum-style diagnostics out of normal hot paths while preserving
  certification and benchmarking routes.
- The benchmark and test methodology tying the above theory to reproducible
  Torch/CUDA-oriented validation.

This notice does not claim ownership of standard, previously known mathematics,
public-domain ideas, or classical formulas themselves, such as standard
eigenvalue decompositions, the classical Hellmann-Feynman identity, Jordan
normal form, Puiseux series, or other known results. The attribution claim is
for Dmitry Sierikov's original selection, naming, synthesis, extensions,
experimental theory, software realization, validation methodology, and the
specific intellectual work expressed in this repository.

Any academic, research, benchmark, documentation, product, or derivative work
that relies on these original SFT/SFT Torch theory contributions should cite
or otherwise visibly credit Dmitry Sierikov and this repository:

   Dmitry Sierikov, Spectral Flow Transform / SFT Torch,
   https://github.com/dimaq12/spectral_flow

Warranty Disclaimer
-------------------

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT, AND TITLE. IN NO EVENT
SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES, OR
OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT, OR OTHERWISE,
ARISING FROM, OUT OF, OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.
