LLMs在隐式推理过程中不是按照逐步思考的。

LLMs Do Not Think Step-by-step In Implicit Reasoning

November 24, 2024

作者: Yijiong Yu

cs.AI

摘要

众所周知，思维链（Chain-of-Thought）可以显著提升大型语言模型（LLMs）在复杂任务上的表现。然而，由于它会导致推理速度变慢和计算成本增加，许多研究尝试使用隐式思维链（implicit CoT），这种方法不需要LLMs明确生成中间步骤。但它们的有效性与典型的显式思维链方法之间仍存在差距。这让我们怀疑，隐式思维链是否真的等同于显式思维链？因此，在这项研究中，我们通过实验来探讨这个问题。当LLMs执行隐式思维链时，我们从模型的隐藏状态中探测中间步骤的信息。结果令人惊讶地表明，LLMs几乎不考虑中间步骤，这表明它们可能只依赖经验而非严格的逐步推理。此外，我们发现LLMs的隐式推理能力易受影响且不稳定，再次证实了显式思维链对于有效支持复杂任务的必要性。

English

It has been well-known that Chain-of-Thought can remarkably enhance LLMs' performance on complex tasks. However, because it also introduces slower inference speeds and higher computational costs, many researches have attempted to use implicit CoT, which does not need LLMs to explicitly generate the intermediate steps. But there is still gap between their efficacy and typical explicit CoT methods. This leaves us a doubt that, does implicit CoT really equal to explicit CoT? Therefore, in this study, we address this question through experiments. We probe the information of intermediate steps from the model's hidden states when it is performing implicit CoT. The results surprisingly indicate that LLMs hardly think about intermediate steps, suggesting they may just rely on experience rather than strict step-by-step reasoning. Moreover, we find LLMs' implicit reasoning capabilities are susceptible and unstable, reaffirming the necessity of explicit CoT to effectively support complex tasks.