Comment by jeswin

For a while now, I've found pelican rendering to be an unreliable metric for LLM ability - and most people know it. Yet, somehow it gets upvoted to the very top of every new model discussion.