Short and Sweet: Enhancing LLM Performance with Constrained Chain-of-Thought | Towards Data Science

Sometimes few words are enough: reducing output length for increasing accuracy