Guide to Model-Based Detection
Model-Based Detection is an automated system that evaluates whether Ethereum addresses are controlled by humans or bots (Sybil entities). It runs in the background, without requiring any active participation from the user.
What is Model-Based Detection?
Model-Based Detection analyzes historical transaction data from an Ethereum address and assigns a score between 0 and 100. The ETH Activity Model specifically examines a user’s transaction history across:
- Ethereum Mainnet (L1)
- Arbitrum
- Base
- Optimism
- Polygon
- zkSync
The resulting score indicates the likelihood that an address is controlled by a real human rather than an automated or Sybil entity.
Impact on Users
- Addresses receive an automatic score based on their on-chain transaction activity.
- That score determines access levels to features, programs, and integrations across the Passport ecosystem.
- Higher scores allow easier access to gated experiences.
- Lower scores may require additional verification before access is granted.
- Users with low Model-Based scores can still earn Passport stamps as an alternative way to prove their humanity.
Benefits
- Enhanced security — helps integrators prevent Sybil attacks at scale.
- Streamlined experience — active, legitimate users typically pass without any extra steps.
- Flexible verification — combines automatic evaluation with stamp-based credentials for users with limited on-chain history.
How it works
| Detail | Value |
|---|---|
| Type | Automatic, address-level evaluation |
| Score range | 0–100 |
| User action required | None |
| Networks analyzed | Ethereum, Arbitrum, Base, Optimism, Polygon, zkSync |
| Triggered by | Interactions with supported dApps |
The evaluation occurs during interactions with supported dApps. Scores then influence feature access across the broader dApp ecosystem.
Developer Resources
Developer tools are built around Model-Based Detection so integrators can incorporate the scores into their own access flows. For implementation details, refer to the Human Passport developer documentation portal .