arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.13419cs.CLcs.AIcs.CYcs.LGcs.SC

从词语到世界:认知架构的组合性

From Words to Worlds: Compositionality for Cognitive Architectures

Ruchira Dhar, Anders Søgaard

AI总结:

该研究针对大语言模型的组合性问题,对4个家族的12个模型和3类任务开展实证分析,发现规模扩大会提升组合能力,但指令微调往往起反作用,为优化符合人类认知的大语言模型提出了待解问题。

摘要:

大语言模型(LLM)是性能极强的联结主义系统,但它们是否展现出更强的组合性?更重要的是,这是否是其表现优异的部分原因?我们对4个LLM家族(共12个模型)和3类任务开展了实证分析,其中包括下文提出的一项新任务。研究结果揭示了LLM学习组合策略的复杂关系——模型规模扩大会提升组合能力,但指令微调往往会产生相反效果。这种差异为开发和优化符合人类认知能力的大语言模型带来了若干待解决问题。

英文摘要:

Large language models (LLMs) are very performant connectionist systems, but do they exhibit more compositionality? More importantly, is that part of why they perform so well? We present empirical analyses across four LLM families (12 models) and three task categories, including a novel task introduced below. Our findings reveal a nuanced relationship in learning of compositional strategies by LLMs -- while scaling enhances compositional abilities, instruction tuning often has a reverse effect. Such disparity brings forth some open issues regarding the development and improvement of large language models in alignment with human cognitive capacities.

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。

↑