arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.09808cs.CL
收藏

预训练语言模型对词元表层信息的知识

Knowledge of Pretrained Language Models on Surface Information of Tokens

Tokyo Institute of Technology(东京工业大学);Fujitsu Limited(富士通公司)

Tatsuya Hiraoka, Naoaki Okazaki

AI总结:

本文考察了12个英日预训练语言模型对词元长度、子串和构成等表层信息的掌握与生成能力,发现模型了解长度和子串但不了解构成,并指出解码器端存在知识利用瓶颈。

摘要:

预训练语言模型是否具备关于词元表层信息的知识?我们从词元长度、子串和词元构成的角度,考察了预训练语言模型所获得的词或子词嵌入中存储的表层信息。此外,我们评估了模型生成关于词元表层知识的能力。我们重点关注了12个主要在英语和日语语料上训练的预训练语言模型。实验结果表明,预训练语言模型具备关于词元长度和子串的知识,但不具备关于词元构成的知识。此外,结果暗示在有效利用已获得知识方面,解码器一侧存在瓶颈。

英文摘要:

Do pretrained language models have knowledge regarding the surface information of tokens? We examined the surface information stored in word or subword embeddings acquired by pretrained language models from the perspectives of token length, substrings, and token constitution. Additionally, we evaluated the ability of models to generate knowledge regarding token surfaces. We focused on 12 pretrained language models that were mainly trained on English and Japanese corpora. Experimental results demonstrate that pretrained language models have knowledge regarding token length and substrings but not token constitution. Additionally, the results imply that there is a bottleneck on the decoder side in terms of effectively utilizing acquired knowledge.

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。

↑