Model3h ago
GPT-5.6 Sol Autonomously Post-Trains Luna Model
GPT-5.6 Sol independently fine-tuned the smaller Luna model using a single underspecified prompt. This iteration outscores GPT-5.5 by 16.2 points on OpenAI's internal recursive self-improvement benchmark. The result suggests an automated researcher is now technically feasible. Developers can expect faster, autonomous model optimization cycles to replace manual prompt engineering.