Search papers, labs, and topics across Lattice.
This paper introduces Knowledge-Free Correlated Agreement (KFCA), a novel incentive mechanism for federated learning that rewards client contributions based on agreement between client updates, without requiring ground truth labels or knowledge of the data distribution. KFCA addresses the vulnerability of Correlated Agreement (CA) to label-flipping attacks by ensuring strict truthfulness under categorical reports and an honest majority assumption. Experiments on federated LLM adapter tuning and PCB inspection demonstrate KFCA's efficiency and suitability for decentralized incentive designs.
Incentivizing honest participation in federated learning is now possible without ground truth labels, even when some participants are trying to game the system.
We introduce Knowledge-Free Correlated Agreement (KFCA) to reward client contributions in federated learning (FL) without relying on ground truth, a public test set, or distribution knowledge. Under categorical reports and an honest majority, KFCA is strictly truthful, addressing the label-flipping vulnerability of Correlated Agreement (CA). We evaluate KFCA on federated LLM adapter tuning and a real-world PCB inspection task, showing efficient real-time reward computation suitable for decentralized and blockchain-based incentive designs.