Skip to main navigation menu Skip to main content Skip to site footer

Vol. 9 (2021)

TACL approved

Continual Learning for Grounded Instruction Generation by Observing Human Following Behavior

Published 2022-01-04

  • Noriyuki Kojima
    • ,
  • Alane Suhr
    • ,
  • Yoav Artzi

Noriyuki Kojima
Cornell Univeristy

Alane Suhr
Cornell Univeristy

Yoav Artzi
Cornell Univeristy

Abstract

We study continual learning for natural language instruction generation, by observing human users' instruction execution. We focus on a collaborative scenario, where the system both acts and delegates tasks to human users using natural language. We compare user execution of generated instructions to the original system intent as an indication to the system's success communicating its intent. We show how to use this signal to improve the system's ability to generate instructions via contextual bandit learning. In interaction with real users, our system demonstrates dramatic improvements in its ability to generate language over time.
Presented at EMNLP 2021 Article at MIT Press