探究语言模型指令微调的多语言校准效应
Investigating the Multilingual Calibration Effects of Language Model Instruction-Tuning
AI总结：
本研究探讨了多语言环境下语言模型指令微调对校准的影响，发现高资源语言SFT数据能显著提升模型置信度，但准确性提升有限，揭示了标准SFT在多语言中的局限性。
摘要：
确保深度学习模型在预测不确定性方面得到良好校准对于维持其可信度和可靠性至关重要，尽管基础模型研究不断取得进展，但大型语言模型（LLMs）与其校准之间的关系仍然是一个开放的研究领域。在本工作中，我们关注了多语言环境下LLM校准中的关键缺口，试图更好地理解数据稀缺性如何可能导致不同的校准效应，以及常用技术如何应用于这些环境。我们对两个多语言基准进行分析，分别覆盖29和42种语言，发现即使在低资源语言中，指令微调高资源语言SFT数据集后，模型置信度可以显著增加。然而，准确性改进是微小或不存在的，导致误校准，突显了标准SFT在多语言中的关键缺陷。此外，我们观察到使用标签平滑技术是一种合理的缓解方法，无需低资源SFT数据，能够维持所有语言更好的校准。总体而言，这强调了在训练和微调LLM时考虑多语言的重要性，以提高其在下游应用中的可靠性和公平性。
英文摘要：
Ensuring that deep learning models are well-calibrated in terms of their predictive uncertainty is essential in maintaining their trustworthiness and reliability, yet despite increasing advances in foundation model research, the relationship between such large language models (LLMs) and their calibration remains an open area of research. In this work, we look at a critical gap in the calibration of LLMs within multilingual settings, in an attempt to better understand how the data scarcity can potentially lead to different calibration effects and how commonly used techniques can apply in these settings. Our analysis on two multilingual benchmarks, over 29 and 42 languages respectively, reveals that even in low-resource languages, model confidence can increase significantly after instruction-tuning on high-resource language SFT datasets. However, improvements in accuracy are marginal or non-existent, resulting in mis-calibration, highlighting a critical shortcoming of standard SFT for multilingual languages. Furthermore, we observe that the use of label smoothing to be a reasonable method alleviate this concern, again without any need for low-resource SFT data, maintaining better calibration across all languages. Overall, this highlights the importance of multilingual considerations for both training and tuning LLMs in order to improve their reliability and fairness in downstream use.
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。
