# Word2vec to behavior: morphology facilitates the grounding of language in machines

## TL;DR

This paper introduces a method for training robots to understand and respond to natural language commands by aligning motor actions with word embeddings, demonstrating improved generalization to new commands influenced by robot structure.

## Contribution

The study presents a novel approach that trains robots to associate language with actions using word2vec embeddings, highlighting the impact of robot morphology on language grounding.

## Key findings

- Robots can respond appropriately to unseen commands after training.
- Mechanical structure influences the alignment between language and motor actions.
- Aligning linguistic and motor similarities facilitates language grounding in robots.

## Abstract

Enabling machines to respond appropriately to natural language commands could greatly expand the number of people to whom they could be of service. Recently, advances in neural network-trained word embeddings have empowered non-embodied text-processing algorithms, and suggest they could be of similar utility for embodied machines. Here we introduce a method that does so by training robots to act similarly to semantically-similar word2vec encoded commands. We show that this enables them to act appropriately, after training, to previously-unheard commands. Finally, we show that inducing such an alignment between motoric and linguistic similarities can be facilitated or hindered by the mechanical structure of the robot. This points to future, large scale methods that find and exploit relationships between action, language, and robot structure.