Ensure AI Quality & Earn Rewards
Join the verification network to maintain trust and quality in decentralized AI while earning substantial rewards
Why Become a Verifier?
Play a crucial role in maintaining AI quality while earning substantial rewards
Verifier Tiers & Requirements
Choose your verification tier based on expertise and commitment level
Standard Verifier
Requirements
- 1,000 MLB tokens minimum stake
- Complete verification training course
- Pass initial verification test (80%+)
- Maintain 90%+ accuracy rate
- Verify at least 10 tasks per month
Responsibilities
- Verify basic inference tasks
- Check output format compliance
- Validate result reasonableness
- Report obvious errors or anomalies
Expert Verifier
Requirements
- 5,000 MLB tokens minimum stake
- Domain expertise in specific AI fields
- 95%+ accuracy rate over 100+ verifications
- Advanced verification certification
- Verify at least 25 tasks per month
Responsibilities
- Verify complex multi-modal tasks
- Perform detailed accuracy assessments
- Validate model performance claims
- Mentor standard verifiers
- Participate in dispute resolution
Master Verifier
Requirements
- 25,000 MLB tokens minimum stake
- Recognized AI/ML expert credentials
- 98%+ accuracy rate over 500+ verifications
- Leadership in verification community
- Available for emergency verifications
Responsibilities
- Verify cutting-edge research tasks
- Lead consensus in disputed cases
- Design verification protocols
- Train and certify other verifiers
- Represent verifiers in governance
Types of Verification Tasks
Different verification tasks suited to various skill levels and time commitments
Output Validation
Verify that task outputs match expected format and quality standards.
Examples:
- Format compliance checking
- Basic sanity tests
- Output completeness verification
Accuracy Assessment
Evaluate the correctness and quality of AI model predictions.
Examples:
- Ground truth comparison
- Statistical accuracy analysis
- Quality scoring
Performance Verification
Validate claimed performance metrics and benchmarking results.
Examples:
- Benchmark reproduction
- Performance claim validation
- Resource usage verification
Dispute Resolution
Resolve conflicts between different verification results.
Examples:
- Consensus arbitration
- Expert review of disputed cases
- Final decision making
🖥️ Verification Interface Walkthrough
Step-by-step guide through the ML Bridge verification interface
Access Verification Dashboard
Navigate to the verification section and review available tasks.
Interface Elements:
Click "Verification" in the main navigation
List of tasks awaiting verification
Your reputation, earnings, and verification history
Stake MLB Tokens
Stake the required amount of MLB tokens to participate in verification.
Interface Elements:
Dedicated staking panel for verifiers
Choose verification tier based on stake amount
Enter MLB tokens to stake (minimum requirements shown)
Select Verification Tasks
Choose tasks to verify based on your expertise and tier.
Interface Elements:
Filter by model type, complexity, and reward amount
Individual task cards with details and requirements
Commit to verifying the selected task