Syntax through rapid synaptic changes
Syntax through rapid synaptic changes
复制标题
DOI：
10.1101/2023.12.21.572018
复制
发表时间：
2023-12
期刊：
影响因子：
--
通讯作者：
Lin Sun;Sanjay G. Manohar
中科院分区：
文献类型：
--
作者：
Lin Sun;Sanjay G. Manohar
神经元是人类语言的核心组织成分，但很少有模型解释它如何在神经元中实现。我们结合了两个快速突触规则来演示神经元如何实现简单的语法。单词绑定到句法角色（例如“狗”作为主语或宾语），并且角色遵循由预定义的句法知识指导的排序规则（例如主语→动词→宾语）。我们发现，与人类一样，该模型比打乱的单词列表更好地回忆句子，并且它可以将单词序列化以表达一个句子的想法。该模型还支持顺序自由语素语言，表现出句法启动，并展示了典型的失语症模式时，损坏。至关重要的是，它使用一种直观的表示来实现这些，其中单词填充角色，允许结构化认知。
Syntax is a central organizing component of human language but few models explain how it may be implemented in neurons. We combined two rapid synaptic rules to demonstrate how neurons can implement a simple grammar. Words bind to syntactic roles (e.g. “dog” as subject or object) and the roles obey ordering rules (e.g. subject → verb → object), guided by predefined syntactic knowledge. We find that, like humans, the model recalls sentences better than shuffled word-lists, and it can serialize words to express an idea as a sentence. The model also supports order-free morphemic languages, exhibits syntactic priming and demonstrates typical patterns of aphasia when damaged. Crucially, it achieves these using an intuitive representation where words fill roles, allowing structured cognition.