What is a t-math-latex input method engine?

I don't seem to be able to find this anywhere in the documentation, and also in the source, but I do have something called t-math-latex, which I don't really understand.

Could this be added to the documentation?