Passport XYZ's Ethereum and NFT Stamps use Model-Based Detection to score you based on how similar your transaction history resembles Sybil or human behavior.
Last updated
Was this helpful?
The Ethereum and NFT Stamps are backed by supervised machine learning algorithms that score your address based on your wallet's transaction history. The Ethereum Stamp examines your activity across Ethereum Mainnet (L1), Arbitrum, Base, Optimism, Polygon, and zkSync, while the NFT Stamp only looks at activity on Ethereum Mainnet (L1).
The higher score you get, the more your transaction history resembles human behavior, and the more points you'll get from each Stamp.
If you did not receive points, it doesn't imply you're seen as a Sybil entity. Rather, it means the model requires more robust activity data to confidently assign points under the new system.
The specifics of the model's criteria are kept under wraps to prevent and other bad actors from being able to manipulate the system. Rest assured, the model is designed to reward authentic, varied, and consistent participation on the network, and have been rated as the best-in-class Sybil detection model in the industry by third-party experts.
The 90-day renewal cycle ensures that the stamp remains a current and accurate portrayal of your latest activities. This regular update aligns with the fast-paced evolution of the Ethereum ecosystem and allows for adjustments that reflect your most recent contributions.
The Model-based Stamps are divided into different groups of credentials. You'll notice that one section awards you based on how robust and human-looking your relevant transaction history appears, while others award you based on specific holdings.
We're periodically refining these Stamps to more accurately reflect the vibrant dynamism of Ethereum transactions. While previous versions had merits, newer versions of the Stamps embraces a broader spectrum of engagements, ensuring a more inclusive and comprehensive representation of community involvement.
For more details or assistance with your model-based Stamps, please .
