arXiv 2601.002822026-08-27cs.CLcs.AIcs.LG

大语言模型还能自我解释吗？探讨量化对自我解释的影响

Can Large Language Models Still Explain Themselves? Investigating the Impact of Quantization on Self-Explanations

Quality and Usability Lab, Technische Universität Berlin（柏林技术大学质量与可用性实验室）；University of Copenhagen（哥本哈根大学）；Saarland Informatics Campus（萨尔州信息学校园）；German Research Center for Artificial Intelligence (DFKI)（德国人工智能研究中心）；Centre for European Research in Trusted AI (CERTAIN)（可信AI欧洲研究中心）；BIFOLD – Berlin Institute for the Foundations of Learning and Data（柏林学习与数据基础研究院）

Qianli Wang, Nils Feldhus, Pepa Atanasova, Fedor Splitt, Simon Ostermann, Sebastian Möller, Vera Schmitt

备注：: EMNLP 2026 Findings; camera-ready version

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

研究探讨量化对大语言模型自我解释质量及可信度的影响，发现量化会导致中等程度的下降，但不影响其作为压缩技术的有效性。

摘要：

量化被广泛用于加速推理并简化大型语言模型（LLMs）的部署，但其对自我解释（SEs）的影响尚无研究。SEs是由LLMs生成以证明其自身输出的解释，需要对模型自身决策过程进行推理，这种能力可能对量化特别敏感。随着SEs在高风险应用中越来越被依赖，理解量化是否以及在多大程度上会降低SE的质量和可信度至关重要。为解决这一差距，我们检查了两种类型的SEs：自然语言解释（NLEs）和反事实示例，这些SEs由使用三种常见技术在不同位宽下量化生成的LLMs生成。我们的发现表明，量化通常会导致SE质量（最高下降4.4%）和可信度（最高下降2.38%）的中等下降。用户研究进一步表明，量化会降低SEs的连贯性和可信度（最高下降8.5%）。与较小的模型相比，较大的模型在SE质量方面对量化有有限的抗性，但在可信度方面保持更好。此外，没有一种量化技术在任务准确性、SE质量和可信度上都表现优异。鉴于量化的影响因上下文而异，我们建议针对特定应用场景验证SE质量，尤其是对于NLEs，它们显示出更大的敏感性。不过，SE质量和可信度的相对轻微下降并不影响量化作为模型压缩技术的有效性。

英文摘要：

Quantization is widely used to accelerate inference and streamline the deployment of large language models (LLMs), yet its effects on self-explanations (SEs) remain unexplored. SEs, generated by LLMs to justify their own outputs, require reasoning about the model's own decision-making process, a capability that may exhibit particular sensitivity to quantization. As SEs are increasingly relied upon for transparency in high-stakes applications, understanding whether and to what extent quantization degrades SE quality and faithfulness is critical. To address this gap, we examine two types of SEs: natural language explanations (NLEs) and counterfactual examples, generated by LLMs quantized using three common techniques at distinct bit widths. Our findings indicate that quantization typically leads to moderate declines in both SE quality (up to 4.4%) and faithfulness (up to 3.9%). The user study further demonstrates that quantization considerably diminishes both the coherence and trustworthiness of SEs (by up to 8.5%). Compared to smaller models, larger models show limited resilience to quantization in terms of SE quality but maintain more faithfulness. Moreover, no quantization technique consistently excels across task accuracy, SE quality, and faithfulness. Because quantization's impact varies considerably by context and can be sizable in specific cases, we recommend validating SE quality for the intended use case. Despite these sometimes considerable drops, quantization remains an effective compression technique when its impact on SEs is properly validated.

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。