大語(yǔ)言模型時(shí)代的術(shù)語(yǔ)翻譯

打開(kāi)文本圖片集

摘 要：大語(yǔ)言模型為術(shù)語(yǔ)學(xué)發(fā)展提供了千載難逢的發(fā)展機(jī)會(huì)，我國(guó)的術(shù)語(yǔ)學(xué)研究應(yīng)當(dāng)緊緊抓住這個(gè)機(jī)會(huì)。在大語(yǔ)言模型時(shí)代，我們應(yīng)當(dāng)把術(shù)語(yǔ)的民族化和術(shù)語(yǔ)的國(guó)際化結(jié)合起來(lái)，從而推動(dòng)我國(guó)術(shù)語(yǔ)學(xué)的發(fā)展。

關(guān)鍵詞：術(shù)語(yǔ)學(xué)；大語(yǔ)言模型；民族化；國(guó)際化

中圖分類號(hào)：H083; G20  文獻(xiàn)標(biāo)識(shí)碼：A  DOI：10.12339/j.issn.1673-8578.2024.03.011

Term Translation in Era of Large Language Models//FENG Zhiwei

Abstract： The Large Language Models provide a golden opportunity for the development of terminology studies， and China’s terminology studies should firmly grasp this opportunity. In the era of Large Language Models，we should combine the nationalization of terminology and the internationalization of terminology to promote the development of terminology in China.

Keywords： terminology; large language model; nationalization; internationalization

收稿日期：2024-01-24  修回日期：2024-03-01

1 GPT系列產(chǎn)品的出現(xiàn)

2018年，OpenAI公司開(kāi)發(fā)了“基于轉(zhuǎn)換器的生成式預(yù)訓(xùn)練模型”，其英文術(shù)語(yǔ)是Generative Pre-trained Transformer，簡(jiǎn)稱GPT-1。（剩余3989字）