Tag
This research paper proposes a framework for fair token allocation and private data valuation in decentralized multi-modal agentic systems, using differentially private prototypes to balance privacy and utility while scheduling limited edge AI resources.