o1 Strawberry Model

While the previous GPT-4o model was confidently misdiagnosing, o1 takes it a step further—not only confidently misdiagnosing but also rationalizing