The Fast and the Flexible: training neural networks to learn to follow instructions from small data

TL;DR

This paper presents a neural network approach that learns to follow human instructions with minimal data by combining offline general learning with fast online adaptation, effectively handling novel vocabulary and diverse language use.

Contribution

It introduces a two-phase training method enabling neural networks to automatically acquire inductive biases for instruction following from small data, without manual feature engineering.

Findings

Efficient adaptation to new vocabulary with familiar instructions.

Effective instruction following even with human language variability.

Neural network learns general task structure in offline phase.

Abstract

Learning to follow human instructions is a long-pursued goal in artificial intelligence. The task becomes particularly challenging if no prior knowledge of the employed language is assumed while relying only on a handful of examples to learn from. Work in the past has relied on hand-coded components or manually engineered features to provide strong inductive biases that make learning in such situations possible. In contrast, here we seek to establish whether this knowledge can be acquired automatically by a neural network system through a two phase training procedure: A (slow) offline learning stage where the network learns about the general structure of the task and a (fast) online adaptation phase where the network learns the language of a new given speaker. Controlled experiments show that when the network is exposed to familiar instructions but containing novel words, the model adapts very efficiently to the new vocabulary. Moreover, even for human speakers whose language usage can depart significantly from our artificial training language, our network can still make use of its automatically acquired inductive bias to learn to follow instructions more effectively.