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 2305.08246cs.CLcs.AIcs.LG
收藏

学习非语言技能而不牺牲语言熟练度

Learning Non-linguistic Skills without Sacrificing Linguistic Proficiency

Virginia Tech(弗吉尼亚理工大学);Stevens Institute of Technology(史蒂文斯理工学院)

Mandar Sharma, Nikhil Muralidhar, Naren Ramakrishnan

AI总结:

本研究针对LLM学习非语言技能时导致语言能力灾难性遗忘的问题,提出基于信息论干预和技能特定损失的新框架,在仅用1/4非语言数据且无需额外语言数据下,实现非语言技能与语言保留双优。

摘要:

近年来,数学-自然语言处理(Math-NLP)领域见证了显著的增长,其动机源于扩展大语言模型(LLM)性能以学习非语言概念(如数字,以及随后的算术推理)的愿望。然而,非语言技能的注入通常会给LLM带来代价:它会导致核心语言技能的灾难性遗忘,这一后果在文献中往往未被解决。由于Math-NLP已经能够创建出能紧密接近小学生数学技能或计算器算术推理技能的LLM,如果这些模型同时丧失其语言能力,其实用性就会失败。在这项工作中,我们更深入地研究了与LLM相关的灾难性遗忘现象,并随后提出了一种基于信息论干预和技能特定损失的新型非语言技能注入框架,使LLM能够学习严格的算术推理。我们的模型在注入的非语言技能和语言知识保留方面均优于现有最先进水平,并且仅使用了非语言训练数据的一小部分(1/4)和零额外的合成语言训练数据。

英文摘要:

The field of Math-NLP has witnessed significant growth in recent years, motivated by the desire to expand LLM performance to the learning of non-linguistic notions (numerals, and subsequently, arithmetic reasoning). However, non-linguistic skill injection typically comes at a cost for LLMs: it leads to catastrophic forgetting of core linguistic skills, a consequence that often remains unaddressed in the literature. As Math-NLP has been able to create LLMs that can closely approximate the mathematical skills of a grade-schooler or the arithmetic reasoning skills of a calculator, the practicality of these models fail if they concomitantly shed their linguistic capabilities. In this work, we take a closer look into the phenomena of catastrophic forgetting as it pertains to LLMs and subsequently offer a novel framework for non-linguistic skill injection for LLMs based on information theoretic interventions and skill-specific losses that enable the learning of strict arithmetic reasoning. Our model outperforms the state-of-the-art both on injected non-linguistic skills and on linguistic knowledge retention, and does so with a fraction of the non-linguistic training data (1/4) and zero additional synthetic linguistic training data.

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。

↑