Despite the recent progress in text summarization made by large language models (LLMs), they often generate summaries that are factually inconsistent with original articles, known as "hallucinations" in text generation. Unlike previous small models (e.g., BART, T5), current LLMs make fewer silly mistakes but more sophisticated ones, such as imposing cause and effect, adding false details, and overgeneralizing, etc. These hallucinations are challenging to detect through traditional methods, which poses great challenges for improving the factual consistency of text summarization. In this paper, we propose an adversarially DEcoupling method to disentangle the Comprehension and EmbellishmeNT abilities of LLMs (DECENT). Furthermore, we adopt a probing-based parameter-efficient technique to cover the shortage of sensitivity for true and false in the training process of LLMs. In this way, LLMs are less confused about embellishing and understanding, thus can execute the instructions more accurately and have enhanced abilities to distinguish hallucinations. Experimental results show that DECENT significantly improves the reliability of text summarization based on LLMs.

翻译：尽管大型语言模型在文本摘要领域取得了近期进展，但它们生成的摘要常与原文存在事实不一致，即文本生成中的“幻觉”现象。与以往的小型模型（如BART、T5）不同，当前LLM较少出现明显错误，但会产生更复杂的错误，例如强加因果关系、添加虚假细节和过度概括等。这些幻觉难以通过传统方法检测，给提升文本摘要的事实一致性带来了巨大挑战。本文提出一种对抗性解耦方法（DECENT），用于分离LLM的理解与润色能力。此外，我们采用基于探针的参数高效技术，弥补LLM训练过程中对真假敏感性不足的问题。通过这种方式，LLM在润色与理解方面的混淆程度降低，从而能更准确地执行指令，并增强区分幻觉的能力。实验结果表明，DECENT显著提升了基于LLM的文本摘要的可靠性。