Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learning
Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learning
复制标题
多智能体通信与自然语言的结合：功能性语言学习和结构性语言学习之间的协同作用
DOI：
--
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
O. Tieleman
中科院分区：
文献类型：
--
作者：
Angeliki Lazaridou;Anna Potapenko;O. Tieleman
我们提出了一种将多智能体通信和传统的数据驱动方法相结合的方法来进行自然语言学习，最终目标是教智能体用自然语言与人类进行通信。我们的出发点是一个语言模型，它是在通用的而不是特定于任务的语言数据上训练的。然后，我们将这个模型放置在一个多智能体的自我游戏环境中，该环境生成用于适应或调节模型的特定任务奖励，将其转化为任务条件语言模型。我们介绍了一种新的方法，结合这两种类型的学习的基础上重新排序的语言模型样本的想法，并表明，这种方法优于其他人在与人类沟通的视觉参考通信任务。最后，我们提出了一个分类的不同类型的语言漂移，可以发生在旁边的一组措施来检测它们。
We present a method for combining multi-agent communication and traditional data-driven approaches to natural language learning, with an end goal of teaching agents to communicate with humans in natural language. Our starting point is a language model that has been trained on generic, not task-specific language data. We then place this model in a multi-agent self-play environment that generates task-specific rewards used to adapt or modulate the model, turning it into a task-conditional language model. We introduce a new way for combining the two types of learning based on the idea of reranking language model samples, and show that this method outperforms others in communicating with humans in a visual referential communication task. Finally, we present a taxonomy of different types of language drift that can occur alongside a set of measures to detect them.