This started as a personal experiment: one person, one framework, ten AI models.

I built a semantic reasoning engine (WFGY: All Principles Return to One) and tested how well each model could handle abstract logic, conceptual shifts, and consistent inference—all using the same PDF.

The results are posted above. No fancy wrappers, no login walls—just raw data, an illustrated battle poster, and the full experiment.

Yes, it's a bit weird. But it's real. And honestly? I just hope someone out there sees the effort and the courage it took to do this solo.

Happy to answer questions. Would love your feedback, criticism, or even memes. Thanks for taking a look