It's also relying on the assumption that the checking LLM only ever corrects wro...

Systemerror7A69 16 days ago | parent | context | favorite | on: How I use LLMs to learn complex topics

It's also relying on the assumption that the checking LLM only ever corrects wrong statements and never incorrectly "corrects" an already correct statement, which might not always be the case as well.