Search papers, labs, and topics across Lattice.
This paper introduces an active auditing framework for Decentralized Federated Learning (DFL) to defend against adaptive backdoor attacks, moving beyond passive defense mechanisms. They model the diffusion of adversarial updates across graph topologies and propose proactive auditing metrics based on private probes to expose latent backdoors. The framework also includes a topology-aware defense placement strategy, demonstrating strong performance against stealthy attacks while maintaining utility.
Active probing reveals backdoors that passive defenses miss in decentralized federated learning.
Decentralized Federated Learning (DFL) remains highly vulnerable to adaptive backdoor attacks designed to bypass traditional passive defense metrics. To address this limitation, we shift the defensive paradigm toward a novel active, interventional auditing framework. First, we establish a dynamical model to characterize the spatiotemporal diffusion of adversarial updates across complex graph topologies. Second, we introduce a suite of proactive auditing metrics, stochastic entropy anomaly, randomized smoothing Kullback-Leibler divergence, and activation kurtosis. These metrics utilize private probes to stress-test local models, effectively exposing latent backdoors that remain invisible to conventional static detection. Furthermore, we implement a topology-aware defense placement strategy to maximize global aggregation resilience. We provide theoretical property for the system's convergence under co-evolving attack and defense dynamics. Numeric empirical evaluations across diverse architectures demonstrate that our active framework is highly competitive with state-of-the-art defenses in mitigating stealthy, adaptive backdoors while preserving primary task utility.
