Key insights on #LLMs & #ChainofThought reasoning: 1️⃣ More steps = Smarter #AI 2️⃣ Length depends on complexity: Short for simple, long for complex 3️⃣ A twist: Lengthy, imperfect reasoning can still be effective! 🔗 Check the paper: lnkd.in/dFVMUrXD #prompting #gpt