更大的语言模型不在乎你怎么想：为什么思维链提示在主观任务中会失败
Larger Language Models Don't Care How You Think: Why Chain-of-Thought Prompting Fails in Subjective Tasks
AI总结：
本文研究了思维链提示在主观任务中失败的原因，发现对于更大的语言模型，CoT与上下文学习一样会因检索固化的推理先验而导致后验坍缩。
摘要：
大型语言模型（LLM）中的上下文学习（ICL）已成为执行自然语言任务的主流技术，因为它不需要通过基于梯度的方法更新模型参数。ICL有望以极低的计算成本“适应”LLM，使其在当前任务上达到具有竞争力或最先进的水平。通过在提示中明确引入得出最终标签的推理过程，可以增强ICL，这种技术称为思维链（CoT）提示。然而，最近的研究发现，ICL主要依赖于任务先验的检索，而不是“学习”执行任务，特别是在情感和道德等复杂的主观领域，先验会使后验预测固化。在这项工作中，我们研究了“启用”推理是否也会在LLM中产生相同的行为，即CoT的格式检索到的推理先验尽管提示中存在证据，但仍保持相对不变。我们发现，令人惊讶的是，对于更大的语言模型，CoT确实和ICL一样存在后验坍缩现象。代码可在https://github.com/gchochla/cot-priors获取。
英文摘要：
In-Context Learning (ICL) in Large Language Models (LLM) has emerged as the dominant technique for performing natural language tasks, as it does not require updating the model parameters with gradient-based methods. ICL promises to "adapt" the LLM to perform the present task at a competitive or state-of-the-art level at a fraction of the computational cost. ICL can be augmented by incorporating the reasoning process to arrive at the final label explicitly in the prompt, a technique called Chain-of-Thought (CoT) prompting. However, recent work has found that ICL relies mostly on the retrieval of task priors and less so on "learning" to perform tasks, especially for complex subjective domains like emotion and morality, where priors ossify posterior predictions. In this work, we examine whether "enabling" reasoning also creates the same behavior in LLMs, wherein the format of CoT retrieves reasoning priors that remain relatively unchanged despite the evidence in the prompt. We find that, surprisingly, CoT indeed suffers from the same posterior collapse as ICL for larger language models. Code is avalaible at https://github.com/gchochla/cot-priors.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
