“ Second, there are no surface form clues to exploit and bias the search towards the ground truth proof” Foundational models are usually trained with a lot of

“ Second, there are no surface form clues to exploit and bias the search towards the ground truth proof” Foundational models are usually trained with a lot of stuff before doing these kinds of tests. Can we know the above statement is true? That something (a) wasn’t in the training data and (b) didn’t have surface-level clues a ML algorithm could spot which the authors didn’t? I felt like asking the latter because both GA’s and NN’s have found simple patterns in problems that humans missed for a long time. They used those patterns to heuristically solve those problems. It might be hard to design tests that eliminate a factor humans can’t see.