arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.09049cs.CLcs.LG
收藏

从大型预训练语言模型中发现有用的句子表示

Discovering Useful Sentence Representations from Large Pretrained Language Models

Scale AI;Arrive

Nishant Subramani, Nivedita Suresh

摘要:

尽管预训练的语言模型作为构建NLP系统的编码器取得了广泛的成功,但它们作为序列生成任务的解码器并不突出。我们探讨的问题,这些模型是否可以适应被用作通用解码器。为了被认为是“通用的”,解码器必须具有针对任何目标句子$s$的隐式表示,使得它可以在以其表示为条件时准确地恢复该句子。对于在大量英语文本上训练的大型基于transformer的语言模型,我们研究了使用标准优化方法是否可以轻松地发现这种表示。我们提出并比较了三种表示注入技术的转换器为基础的模型和三个相应的方法,映射句子,并从这个表示空间。实验表明,不仅各种流派的句子都存在表示。更重要的是,在不需要复杂的优化算法的情况下,我们的方法几乎可以完美地恢复这些句子,而根本不需要微调底层的语言模型。

英文摘要:

Despite the extensive success of pretrained language models as encoders for building NLP systems, they haven't seen prominence as decoders for sequence generation tasks. We explore the question of whether these models can be adapted to be used as universal decoders. To be considered "universal," a decoder must have an implicit representation for any target sentence $s$, such that it can recover that sentence exactly when conditioned on its representation. For large transformer-based language models trained on vast amounts of English text, we investigate whether such representations can be easily discovered using standard optimization methods. We present and compare three representation injection techniques for transformer-based models and three accompanying methods which map sentences to and from this representation space. Experiments show that not only do representations exist for sentences from a variety of genres. More importantly, without needing complex optimization algorithms, our methods recover these sentences almost perfectly without fine-tuning the underlying language model at all.

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。

↑