arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.00858cs.CL
收藏

大语言模型能理解语境吗?

Can Large Language Models Understand Context?

Georgetown University(乔治城大学);Apple(苹果公司)

Yilun Zhu, Joel Ruben Antony Moniz, Shruti Bhargava, Jiarui Lu, Dhivya Piraviperumal, Site Li, Yuan Zhang, Hong Yu, Bo-Hsiang Tseng

AI总结:

该研究构建了含4项任务9个数据集的语境理解基准,评估发现预训练稠密模型语境理解弱于微调模型,3位训练后量化会不同程度降低LLM的语境理解性能。

摘要:

理解语境是理解人类语言的关键,大语言模型(LLM)已越来越多地展现出令人印象深刻的这一能力。然而,尽管对LLM的评估涵盖了自然语言处理领域的多个方向,但针对其理解语境特征的语言能力的探究仍受到有限关注。本文通过适配现有数据集以适用于生成式模型评估,提出了一个语境理解基准。该基准包含4项不同任务和9个数据集,所有任务都设计了专门的提示词,用以评估模型的语境理解能力。首先,我们评估了LLM在上下文学习预训练场景下的性能。实验结果表明,与最先进的微调模型相比,预训练稠密模型在理解更细微的语境特征方面存在困难。其次,由于LLM压缩在研究和实际应用中的重要性日益提升,我们评估了量化模型在上下文学习设置下的语境理解能力。我们发现,3位训练后量化会导致模型在我们的基准上出现不同程度的性能下降。我们对这些场景进行了广泛分析,以佐证我们的实验结果。

英文摘要:

Understanding context is key to understanding human language, an ability which Large Language Models (LLMs) have been increasingly seen to demonstrate to an impressive extent. However, though the evaluation of LLMs encompasses various domains within the realm of Natural Language Processing, limited attention has been paid to probing their linguistic capability of understanding contextual features. This paper introduces a context understanding benchmark by adapting existing datasets to suit the evaluation of generative models. This benchmark comprises of four distinct tasks and nine datasets, all featuring prompts designed to assess the models' ability to understand context. First, we evaluate the performance of LLMs under the in-context learning pretraining scenario. Experimental results indicate that pre-trained dense models struggle with understanding more nuanced contextual features when compared to state-of-the-art fine-tuned models. Second, as LLM compression holds growing significance in both research and real-world applications, we assess the context understanding of quantized models under in-context-learning settings. We find that 3-bit post-training quantization leads to varying degrees of performance reduction on our benchmark. We conduct an extensive analysis of these scenarios to substantiate our experimental results.

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。

↑