There's an interesting scheme described here: https://www.youtube.com/watch?v=XZJc1p6RE78 https://www.youtube.com/watch?v=XZJc1p6RE78 Essentially by biasing t

There's an interesting scheme described here: https://www.youtube.com/watch?v=XZJc1p6RE78 https://www.youtube.com/watch?v=XZJc1p6RE78 Essentially by biasing the LLM's token choices slightly you can infer with high probability later that the text was generated by it.