Towards Personalized Social Robots: Adaptive Prompting for Real-Time Context-Aware Conversations
Himanshi Lalwani, Hanan Salam
AI summary
Problem
Current LLM-driven social robots lack real-time linguistic personalization, often relying on static prompts or extensive model fine-tuning that fails to adapt dynamically to individual user preferences during live interactions.
Approach
The authors propose adaptive prompting, a framework that structures interactions as a chain of interdependent prompts that continuously update a user profile and use a rolling window of recent dialogue to generate tailored responses without additional training.
Key results
- Formal adaptive prompting framework for real-time linguistic personalization without fine-tuning
- Scalable architecture reducing context overhead while maintaining user adaptation continuity
- Robotic implementation on QTrobot for personalized productivity coaching
- Empirical validation showing improved engagement and coaching effectiveness over non-personalized baselines
Why it matters
Offers a lightweight, training-free personalization method that advances effective, adaptable human-robot interaction in education, therapy, and coaching.
Abstract
Social robots have demonstrated great potential in various domains. Recent advancements in Large Language Models (LLMs) have expanded the conversational capabilities of these robots, enabling more personalized user interactions. However, current systems primarily focus on behavior or task personalization, or they require extensive pre-training and fine-tuning to achieve language personalization. This paper introduces adaptive prompting, a formal framework for real- time linguistic personalization in LLM-driven robots. By struc- turing interaction as a sequence of interdependent prompts, adaptive prompting enables controllable, efficient, and scal- able personalization without additional model training. To validate our approach, we present a system that integrates adaptive prompting in a social robot to dynamically adapt to user attributes and preferences to provide personalized productivity coaching for college students with Attention Deficit Hyperactivity Disorder (ADHD). Our findings demonstrate that personalized coaching via adaptive prompting improves user engagement and overall coaching effectiveness compared to non-personalized coaching. This indicates the effectiveness of the proposed approach for user adaptation and personalization in social robots, particularly in the aforementioned contexts.