[FR] Reflective latent variables

[tailcalled]

It would be nice to have latent variables, as this could make the markets more informative and usable for high-level abstract questions. I wrote a basic description of latent variables here. I am open to providing help with latent variable math and advice about the best latent variable models to use, including further clarification on how they work. I would even be tempted to try to implement a PR for Manifold to add latent variables, but I am not familiar with the code base, so it would probably take more time than I have; but I can probably write some code that handles the math for the core scoring of latent variable predictions, if desired.

[tailcalled]

Here is a mockup in case that makes it clearer: https://files.onlinetests.me/lvar.html

[tailcalled]

Latent variables for prediction markets: motivation, technical guide, and design considerations