• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Human System Interactions, 2009. HSI '09 >Generalizing the mirror-neuron-model for thinking processes
【24h】

Generalizing the mirror-neuron-model for thinking processes

机译：泛化思维过程的镜像神经元模型

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Recently it was shown that a unique permanent network for words exists in the brain. Here words are represented by (mirror-) neurons and interconnected through nerves. However it is an open question how the nervous connections should be chosen so that meaningful sentences arise. Neurons situated far above the word level must be responsible for this purpose, both for generating and for understanding sentences. In order to find the crucial organization of these neurons it is necessary to determine what constitutes the core of information of a sentence at first. According to a definition provided by the philosopher L. Wittgenstein this core is called a ldquothoughtrdquo. In a pioneering new system consisting of several hierarchically organized networks, each of optimum structure, the concentrated information called ldquothoughtrdquo can be obtained by means of very simple abstraction processes and at last it can be represented by a single neuron in its special network surroundings. This can be regarded as a complex or generalized neural mirror process between a sentence and a thought neuron. In addition, in higher levels of a hierarchical system these neurons can be associatively connected one after another only through nerves. In particular by means of that method, a nearly unlimited amount of different natural language texts can be produced. The described mechanism leads to a completely new paradigm about ldquothinkingrdquo that is very different to the ideas in the world of computer programs. The new theory is based upon the assumption that neurons do not store or process coded information but instead they only represent information. After all, these representatives of information - simple neural elements without symbols - can be traced back to original language-elements like words, sentences or even long texts originating at the outside of the brain. Tests with large coherent texts of well known poets and writers showed that the system worked correctly without any -nexplicit grammar rules. It relied entirely on specific network structures.
机译：最近显示，大脑中存在一个独特的单词永久网络。在这里，单词由（镜像）神经元表示，并通过神经相互连接。然而，如何选择神经连接以使有意义的句子出现是一个悬而未决的问题。远远高于单词级别的神经元必须为此目的负责，既要生成又要理解句子。为了找到这些神经元的关键组织，有必要首先确定什么构成句子信息的核心。根据哲学家L.维特根斯坦（L. Wittgenstein）提供的定义，该核心称为ldquothoughtrdquo。在一个由几个分层组织的网络组成的具有开拓性的新系统中，每个网络都具有最佳结构，可以通过非常简单的抽象过程获得称为“集中信息”的集中信息，最后可以在其特殊网络环境中用单个神经元表示该信息。这可以看作是句子和思想神经元之间的复杂或广义的神经镜像过程。此外，在更高层次的系统中，这些神经元只能通过神经相互关联。特别是通过该方法，可以产生几乎无限量的不同自然语言文本。所描述的机制导致了关于“ ldinghinkingrdquo”的全新范例，该范例与计算机程序世界中的思想非常不同。新理论基于这样的假设，即神经元不存储或处理编码信息，而仅代表信息。毕竟，这些信息的代表-没有符号的简单神经元素-可以追溯到原始的语言元素，例如单词，句子甚至是源自大脑外部的长文本。对著名诗人和作家的大量连贯文章的测试表明，该系统可以正常运行，而无需任何-nexplicit语法规则。它完全依赖于特定的网络结构。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号