While DeepSeek LLMs have shown spectacular abilities, they don't seem to be without their limits. Here are some probable negatives of these products: Not one of the GPT-4o or Claude 3.5 Sonnets could remedy this straightforward problem properly. Only o1 was capable of finding the proper reply with none help. https://x.com/kidtsang/status/1884008035535782292