2024-12-21T14:30:04.397747 | 🔗
With o3 and ARC, we have shown that people can build LLM system to fit with an eval of machine-gradeable answers.