ELITE: Experiential Learning and Intent-Aware Transfer for Self-improving Embodied Agents

Researchers introduce ELITE, a new framework that enables AI embodied agents to learn from their own experiences and transfer knowledge to similar tasks. The system addresses failures in vision-language models when performing complex physical tasks by using self-reflective knowledge construction and intent-aware retrieval mechanisms.

• →ELITE framework enables embodied AI agents to continuously learn from environment interactions and transfer knowledge to similar tasks.
• →The system addresses critical gaps between static VLM training data and dynamic physical interaction requirements.
• →ELITE achieved 9% and 5% performance improvements over base VLMs on EB-ALFRED and EB-Habitat benchmarks without supervision.
• →The framework uses self-reflective knowledge construction to extract reusable strategies and intent-aware retrieval for task application.
• →Results demonstrate effective generalization to unseen task categories, outperforming state-of-the-art training-based methods.