Can TsT transformations produce genuinely new predictions that distinguish them from mere coordinate relabeling?

Your question asks whether T-duality–S-duality–T-duality (TsT) transformations in theoretical physics produce genuinely new physical predictions, or whether they merely relabel coordinates without changing observable outcomes.

Unfortunately, none of the preprints in this batch directly address TsT transformations, dualities in string theory, or how to distinguish coordinate relabelings from physical predictions.

The archive snippets provided cover transformer neural networks, quantum machine learning, Bell inequalities, and quiver mutations—interesting mathematical physics territory, but orthogonal to your question about duality symmetries and their observational content.

To get a proper answer, you'd want papers that discuss:

• String theory dualities and their physical consequences
• How to extract coordinate-independent predictions from dual formulations
• Specific phenomenological tests that distinguish TsT-related physics from mere redundancy

Would you like to rephrase your question about a topic covered in the Antigravity archive, or ask about how dualities appear in the contexts these preprints do address?