And while it might be able to check whether some calculations are right, because LLMs are not reasoning machines, I am not certain it will be able to check that the statistical methods are appropriate
It can flag things but you wouldn’t want to try to use it as “set and forget” because it produces garbage. It has no actual qualitative ability and it resorts to very stock patterns of response. It literally has about 4 approaches and most are ego gratification for the user.
Comments
https://teche.mq.edu.au/2023/02/why-does-chatgpt-generate-fake-references/
And while it might be able to check whether some calculations are right, because LLMs are not reasoning machines, I am not certain it will be able to check that the statistical methods are appropriate