Analysis of pseudo-random sequences by AI

Are there any papers on studying pseudo-random numbers using AI?

The first idea would be to train an AI to get a initial segment of a pseudo-random sequence of numbers (a binary sequence will probably do it) and the AI ​​returns a probability distribution for the next number. The pseudo-random sequence can be generated in any way. Theoretically, it could also be completely random. In this case, the AI ​​should recognize this and return the uniform distribution (on {0,1}). The intersting case would be if the pseudo-random sequence is given by some algorithm. Here it would then be intersting which types of algorithms the AI ​​can handle and which not.