arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.05958cs.CLcs.AI
收藏

并非所有模型都将语言知识定位在同一个地方--BERToids表示的分层探索

Not All Models Localize Linguistic Knowledge in the Same Place: A Layer-wise Probing on BERToids' Representations

Iran University of Science and Technology(伊朗科技大学);Tehran Institute for Advanced Studies, Khatam University(哈塔姆大学德黑兰高等研究院);University of Tehran(德黑兰大学)

Mohsen Fayyaz, Ehsan Aghazadeh, Ali Modarressi, Hosein Mohebbi, Mohammad Taher Pilehvar

摘要:

大多数关于探测表征的最近的工作都集中在BERT上,假设结果可能与其他模型相似。在这项工作中,我们将探索研究扩展到家庭中的其他两个模型,即ELECTRA和XLNet,表明预训练目标或架构选择的变化可能导致在表征中编码语言信息的不同行为。最值得注意的是,我们观察到ELECTRA倾向于在更深的层中编码语言知识,而XLNet则将其集中在早期层中。此外,前者的模型在微调过程中会发生轻微的变化,而后者则会经历重大的调整。此外,我们表明,基于权重混合评估策略得出结论-这是广泛使用的上下文中逐层探测-可以误导给定跨不同层的表示的规范差距。相反,我们采用了一种替代的信息理论探测与最小的描述长度,最近已被证明提供更可靠和翔实的结果。

英文摘要:

Most of the recent works on probing representations have focused on BERT, with the presumption that the findings might be similar to the other models. In this work, we extend the probing studies to two other models in the family, namely ELECTRA and XLNet, showing that variations in the pre-training objectives or architectural choices can result in different behaviors in encoding linguistic information in the representations. Most notably, we observe that ELECTRA tends to encode linguistic knowledge in the deeper layers, whereas XLNet instead concentrates that in the earlier layers. Also, the former model undergoes a slight change during fine-tuning, whereas the latter experiences significant adjustments. Moreover, we show that drawing conclusions based on the weight mixing evaluation strategy -- which is widely used in the context of layer-wise probing -- can be misleading given the norm disparity of the representations across different layers. Instead, we adopt an alternative information-theoretic probing with minimum description length, which has recently been proven to provide more reliable and informative 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。

↑