This research examines the polycentric governance of digital assets in
Decentralized Autonomous Organizations (DAOs). It offers a theoretical
framework and addresses a critical challenge facing decentralized governance by
developing a method to identify sybils, or spurious identities. The method uses
graph deep learning techniques to identify sybil activity in a DAO governance
dataset (snapshot.org). Specifically, a Graph Convolutional Neural Network
(GCNN) learned voting behaviours and a fast k-means vector clustering algorithm
(FAISS) used the high dimensional embeddings to identify similar nodes in a
graph. The results reveal that deep learning can effectively identify sybils,
reducing the voting graph by 2-5%. This research underscores the importance of
sybil resistance in DAOs and offers a novel perspective on decentralized
governance, informing future policy, regulation, and governance practices.

该研究探讨了去中心化自治组织（DAO）中数字资产的多中心治理，并提供了一个理论框架，通过开发一种识别虚假身份（Sybils）的方法，解决了去中心化治理面临的关键挑战。该方法使用图深度学习技术在一个 DAO 治理数据集（snapshot.org）中识别虚假活动。具体而言，图卷积神经网络（GCNN）学习了投票行为，快速 k 均值向量聚类算法（FAISS）利用高维嵌入识别图中的相似节点。结果表明，深度学习可以有效减少 2-5% 的投票图中的虚假节点。该研究强调了去中心化自治组织中防止虚假身份的重要性，并为去中心化治理提供了一种新的视角，为未来的政策、监管和治理实践提供了指导。

匿名投票网络上的图深度学习用于在多中心治理中识别 Sybil

New Online Communities: Graph Deep Learning on Anonymous Voting Networks  to Identify Sybils in Polycentric Governance

Attention-Aware Social Choice tackles the fundamental conflict faced by some
agent communities between their desire to include all members in the decision
making processes and the limited time and attention that are at the disposal of
the community members. Here, we investigate a combination of two techniques for
attention-aware social choice, namely Natural Language Processing (NLP) and
Sampling. Essentially, we propose a system in which each governance proposal to
change the status quo is first sent to a trained NLP model that estimates the
probability that the proposal would pass if all community members directly vote
on it; then, based on such an estimation, a population sample of a certain size
is being selected and the proposal is decided upon by taking the sample
majority. We develop several concrete algorithms following the scheme described
above and evaluate them using various data, including such from several
Decentralized Autonomous Organizations (DAOs).

使用自然语言处理和抽样相结合的系统，先利用 NLP 模型估计社区成员直接投票时提案通过的概率，再基于此估计选择一定规模的人口样本，通过取样多数决定提案。通过应用到多个去中心化自治组织的数据进行评估，解决一些代理社区在包括所有成员在决策过程中及社区成员有限时间和注意力之间的冲突。