

I’m anti-general purpose LLMs, but this is a valid use case for “ai”. There are known constraints, the model is trained on a very specific dataset, and you have have human experts in the loop for assurance.
This also isn’t new. Protein folding models have been successfully used, and validated, for a number of years.
You could put it up on ResearchGate first to get eyes across it and potentially get an endorsement. Then reach out to the endorser to get sponsored for arXiv.
The other option is to submit it for peer review. You’d need to to look up the best pathway based on your field.