指令微调是否让大语言模型更加一致？
Does Instruction Tuning Make LLMs More Consistent?
AI总结：
本文通过对比10个指令微调LLaMA模型与原始LLaMA-7b，发现指令微调普遍提升了模型对输入扰动的一致性，并从事实回忆的机制角度解释了这一改进。
摘要：
指令微调的目的是实现零样本性能，但已有研究表明指令微调也能改善思维链推理和价值对齐（Si et al., 2023）。本文研究指令微调对“一致性”的影响，即语言模型对输入中微小扰动的敏感程度。我们将10个经过指令微调的LLaMA模型与原始LLaMA-7b模型进行比较，结果表明，几乎所有经过指令微调的模型都变得更加一致，无论是在表示层面还是在零样本及下游任务中的预测层面。我们通过对事实回忆的机制分析来解释这些改进。
英文摘要：
The purpose of instruction tuning is enabling zero-shot performance, but instruction tuning has also been shown to improve chain-of-thought reasoning and value alignment (Si et al., 2023). Here we consider the impact on $\textit{consistency}$, i.e., the sensitivity of language models to small perturbations in the input. We compare 10 instruction-tuned LLaMA models to the original LLaMA-7b model and show that almost across-the-board they become more consistent, both in terms of their representations and their predictions in zero-shot and downstream tasks. We explain these improvements through mechanistic analyses of factual recall.
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。
