跳到主要导航 跳到搜索 跳到主要内容

TRELM: Towards Robust and Efficient Pre-training for Knowledge-Enhanced Language Models

  • Junbing Yan
  • , Chengyu Wang*
  • , Taolin Zhang
  • , Xiaofeng He
  • , Jun Huang
  • , Longtao Huang
  • , Hui Xue
  • , Wei Zhang*
  • *此作品的通讯作者
  • East China Normal University
  • Alibaba Group Holding Ltd.

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

KEPLMs are pre-trained models that utilize external knowledge to enhance language understanding. Previous language models facilitated knowledge acquisition by incorporating knowledge-related pre-training tasks learned from relation triples in knowledge graphs. However, these models do not prioritize learning embeddings for entity-related tokens. Moreover, updating the entire set of parameters in KEPLMs is computationally demanding. This paper introduces TRELM, a Robust and Efficient Pre-training framework for Knowledge-Enhanced Language Models. We observe that entities in text corpora usually follow the long-tail distribution, where the representations of some entities are suboptimally optimized and hinder the pre-training process for KEPLMs. To tackle this, we employ a robust approach to inject knowledge triples and employ a knowledge-augmented memory bank to capture valuable information. Furthermore, updating a small subset of neurons in the feed-forward networks (FFNs) that store factual knowledge is both sufficient and efficient. Specifically, we utilize dynamic knowledge routing to identify knowledge paths in FFNs and selectively update parameters during pre-training. Experimental results show that TRELM reduces pre-training time by at least 50% and outperforms other KEPLMs in knowledge probing tasks and multiple knowledge-aware language understanding tasks.

源语言英语
主期刊名2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, LREC-COLING 2024 - Main Conference Proceedings
编辑Nicoletta Calzolari, Min-Yen Kan, Veronique Hoste, Alessandro Lenci, Sakriani Sakti, Nianwen Xue
出版商European Language Resources Association (ELRA)
16790-16801
页数12
ISBN(电子版)9782493814104
出版状态已出版 - 2024
活动Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024 - Hybrid, Torino, 意大利
期限: 20 5月 202425 5月 2024

出版系列

姓名2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, LREC-COLING 2024 - Main Conference Proceedings

会议

会议Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024
国家/地区意大利
Hybrid, Torino
时期20/05/2425/05/24

指纹

探究 'TRELM: Towards Robust and Efficient Pre-training for Knowledge-Enhanced Language Models' 的科研主题。它们共同构成独一无二的指纹。

引用此