Pedagogical Demonstrations and Pragmatic Learning in Artificial Tutor-Learner Interactions

TL;DR

This paper explores how incorporating pedagogical and pragmatic reasoning into artificial tutor-learner interactions significantly improves learning efficiency by mimicking human teaching and learning behaviors.

Contribution

It introduces mechanisms for pedagogical demonstration and pragmatic inference in artificial agents, enhancing learning from demonstrations in multi-goal environments.

Findings

Substantial improvement over standard learning from demonstrations

Effective implementation of pedagogical and pragmatic mechanisms in artificial agents

Enhanced goal disambiguation through pedagogical demonstrations

Abstract

When demonstrating a task, human tutors pedagogically modify their behavior by either "showing" the task rather than just "doing" it (exaggerating on relevant parts of the demonstration) or by giving demonstrations that best disambiguate the communicated goal. Analogously, human learners pragmatically infer the communicative intent of the tutor: they interpret what the tutor is trying to teach them and deduce relevant information for learning. Without such mechanisms, traditional Learning from Demonstration (LfD) algorithms will consider such demonstrations as sub-optimal. In this paper, we investigate the implementation of such mechanisms in a tutor-learner setup where both participants are artificial agents in an environment with multiple goals. Using pedagogy from the tutor and pragmatism from the learner, we show substantial improvements over standard learning from demonstrations.