Hacker News new | ask | show | jobs
by zlatkov 226 days ago
I haven’t come across any research showing that a specific LLM consistently outperforms others for this. It generally works best with strong reasoning models that produce consistent outputs.