研究人员开发了 Giskard，一种旨在提高大规模去中心化学习的安全性与效率的新协议。Giskard 解决了同时维护数据机密性并防御拜占庭（恶意或故障）参与者所带来的挑战。该协议将参与者组织成一个委员会树，与需要全员通信或给少数节点带来沉重负担的现有方法相比，实现了更具可扩展性的聚合方法。对多达一百万参与者的实验表明，即使有相当比例的拜占庭方，Giskard 也能在保持模型效用的同时降低通信复杂度。 AI
影响 提高了去中心化人工智能模型训练的可扩展性和安全性。
排序理由 该集群包含一篇详细介绍去中心化学习新协议的研究论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2606.19129v1 Announce Type: cross Abstract: Dealing simultaneously with confidentiality and Byzantine behaviors in decentralized learning is a challenging problem. Indeed, in decentralized learning, clients train a machine learning model while keeping their data locally and…
Dealing simultaneously with confidentiality and Byzantine behaviors in decentralized learning is a challenging problem. Indeed, in decentralized learning, clients train a machine learning model while keeping their data locally and share their model parameters or gradients with a …
Dealing simultaneously with confidentiality and Byzantine behaviors in decentralized learning is a challenging problem. Indeed, in decentralized learning, clients train a machine learning model while keeping their data locally and share their model parameters or gradients with a …