ToM as XAI for Human-Robot Interaction
Published:Dec 29, 2025 14:09
•1 min read
•ArXiv
Analysis
This paper proposes a novel perspective on Theory of Mind (ToM) in Human-Robot Interaction (HRI) by framing it as a form of Explainable AI (XAI). It highlights the importance of user-centered explanations and addresses a critical gap in current ToM applications, which often lack alignment between explanations and the robot's internal reasoning. The integration of ToM within XAI frameworks is presented as a way to prioritize user needs and improve the interpretability and predictability of robot actions.
Key Takeaways
- •Proposes viewing Theory of Mind (ToM) in Human-Robot Interaction (HRI) as a form of Explainable AI (XAI).
- •Identifies a gap in current ToM applications regarding the alignment of explanations with internal reasoning.
- •Advocates for integrating ToM within XAI frameworks to prioritize user-centered explanations and improve robot interpretability.
Reference
“The paper argues for a shift in perspective, prioritizing the user's informational needs and perspective by incorporating ToM within XAI.”