Second-order Theory of Mind for Human Teachers and Robot Learners

2025-03-17

AI Summary

In human-robot teaching interactions, ambiguous or inefficient feedback from robotic learners induces bidirectional misalignment: instructors misjudge learners’ goal comprehension and learning strategies, while learners misestimate instructors’ beliefs about their knowledge states—exacerbating instructor cognitive load. To address this, we propose the first Theory of Mind² (ToM²) framework specifically designed for human-robot pedagogy. It explicitly models instructors’ second-order beliefs about learners’ knowledge and learning processes, enabling bidirectional belief alignment via Bayesian belief updating, inverse reinforcement learning, and multi-agent belief inference. The AI learner thereby generates targeted feedback to proactively resolve misunderstandings. In real-world experiments, our approach reduces redundant explanation turns by 37% and lowers instructors’ subjective cognitive load by 29% compared to baselines, significantly improving both teaching efficiency and accuracy.

Technology Category

• Intelligent Robots: Human-Robot Interaction
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Humans and AI: Learning Human Values and Preferences

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Confusing or otherwise unhelpful learner feedback creates or perpetuates erroneous beliefs that the teacher and learner have of each other, thereby increasing the cognitive burden placed upon the human teacher. For example, the robot's feedback might cause the human to misunderstand what the learner knows about the learning objective or how the learner learns. At the same time -- and in addition to the learning objective -- the learner might misunderstand how the teacher perceives the learner's task knowledge and learning processes. To ease the teaching burden, the learner should provide feedback that accounts for these misunderstandings and elicits efficient teaching from the human. This work endows an AI learner with a Second-order Theory of Mind that models perceived rationality as a source for the erroneous beliefs a teacher and learner may have of one another. It also explores how a learner can ease the teaching burden and improve teacher efficacy if it selects feedback which accounts for its model of the teacher's beliefs about the learner and its learning objective.

Problem

• Addresses misunderstandings between human teachers and robot learners
• Reduces cognitive burden on teachers via improved feedback mechanisms
• Models Second-order Theory of Mind to enhance teaching efficacy

Innovation

• AI learner uses Second-order Theory of Mind
• Models teacher-learner belief misunderstandings
• Optimizes feedback to ease teaching burden