DICE: Data Influence Cascade in Decentralized Learning
This work addresses the problem of encouraging participation in decentralized learning for distributed compute networks, offering a novel approach to fair attribution, though it appears incremental as it builds on existing decentralized learning frameworks.
The paper tackles the challenge of fair incentive mechanisms in decentralized learning by attributing contributions to participating nodes, proposing the first method to estimate Data Influence Cascade (DICE) in such environments, which theoretically approximates influence over arbitrary neighbor hops and enables applications like selecting collaborators and detecting malicious behaviors.
Decentralized learning offers a promising approach to crowdsource data consumptions and computational workloads across geographically distributed compute interconnected through peer-to-peer networks, accommodating the exponentially increasing demands. However, proper incentives are still in absence, considerably discouraging participation. Our vision is that a fair incentive mechanism relies on fair attribution of contributions to participating nodes, which faces non-trivial challenges arising from the localized connections making influence ``cascade'' in a decentralized network. To overcome this, we design the first method to estimate \textbf{D}ata \textbf{I}nfluence \textbf{C}ascad\textbf{E} (DICE) in a decentralized environment. Theoretically, the framework derives tractable approximations of influence cascade over arbitrary neighbor hops, suggesting the influence cascade is determined by an interplay of data, communication topology, and the curvature of loss landscape. DICE also lays the foundations for applications including selecting suitable collaborators and identifying malicious behaviors. Project page is available at https://raiden-zhu.github.io/blog/2025/DICE/.