arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.01346cs.CLcs.IR
收藏

机器阅读理解的BERT微调研究

A Pairwise Probe for Understanding BERT Fine-Tuning on Machine Reading Comprehension

Peking University(北京大学);University of Technology Sydney(悉尼科技大学)

Jie Cai, Zhengzhou Zhu, Ping Nie, Qian Liu

摘要:

预训练模型为许多NLP任务带来了显着的改进,并得到了广泛的分析。但人们对微调对特定任务的影响知之甚少。直觉上,人们可能会同意预先训练的模型已经学习了单词的语义表示(例如同义词彼此更接近),并且微调进一步提高了其需要更复杂推理的能力(例如共指消解,实体边界检测等)。然而,如何验证这些论点的分析和定量是一个具有挑战性的任务,有很少的工作集中在这一主题。在本文中,大多数探测任务涉及识别匹配的短语对(例如,共指需要匹配一个实体和代词)的观察的启发,我们提出了一个成对的探测理解BERT微调的机器阅读理解(MRC)任务。具体来说,我们确定了MRC中的五种现象。根据成对探测任务,我们比较了预训练和微调BERT的每一层的隐藏表示的性能。所提出的成对探针消除了不准确的模型训练造成的分心问题,并进行了稳健和定量的比较。我们的实验分析得出了非常有说服力的结论:(1)微调对基本和低层次信息以及一般语义任务的影响不大。(2)对于下游任务所需的特定能力,微调的BERT优于预先训练的BERT,并且这种差距在第五层之后是明显的。

英文摘要:

Pre-trained models have brought significant improvements to many NLP tasks and have been extensively analyzed. But little is known about the effect of fine-tuning on specific tasks. Intuitively, people may agree that a pre-trained model already learns semantic representations of words (e.g. synonyms are closer to each other) and fine-tuning further improves its capabilities which require more complicated reasoning (e.g. coreference resolution, entity boundary detection, etc). However, how to verify these arguments analytically and quantitatively is a challenging task and there are few works focus on this topic. In this paper, inspired by the observation that most probing tasks involve identifying matched pairs of phrases (e.g. coreference requires matching an entity and a pronoun), we propose a pairwise probe to understand BERT fine-tuning on the machine reading comprehension (MRC) task. Specifically, we identify five phenomena in MRC. According to pairwise probing tasks, we compare the performance of each layer's hidden representation of pre-trained and fine-tuned BERT. The proposed pairwise probe alleviates the problem of distraction from inaccurate model training and makes a robust and quantitative comparison. Our experimental analysis leads to highly confident conclusions: (1) Fine-tuning has little effect on the fundamental and low-level information and general semantic tasks. (2) For specific abilities required for downstream tasks, fine-tuned BERT is better than pre-trained BERT and such gaps are obvious after the fifth layer.

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。

↑