[riku_iki]

and how can you tell they reason and not parrot some text in training data?

There are papers about trying LLMs on generated reasoning problems, and they usually fail.

[nuancebydefault]

>Usually

That implies - sometimes not. Which would prove at least some reasoning capabilities.

[riku_iki]

In this case I used 'usually' because don't remember all details and didn't want to generalize by saying 'always', but also training/benchmarking protocol can be flawed, for example LLM still can solve shallow reasoning problem by memorizing pattern.