Evidence Hub
Live audit trail for Bittensor subnet testnet evaluation. Each round maps to the judging criteria below — expand any round to inspect scoring data, critic pool activity, weight assignments, and lifecycle events.
Functional Implementation
End-to-end operational subnet on testnet — deployment, miner/validator logic, task-to-reward flow, network stability.
Evidenced by: set_weights on-chain block, phase transitions, submission counts, lifecycle audit events, subnet owner reward.
Incentive Mechanism Integrity
High-quality miners earn more; spam/low-quality submissions penalized; validators incentivized for honest evaluation.
Evidenced by: quality threshold (75/100) enforcement, per-miner score ranking, zero-weight exclusions, weight distribution, winner selection.
Proof of Intelligence / Effort
Meaningful non-trivial work. Objective evaluation metrics, reproducible scoring, measurable competence signals.
Evidenced by: 10-bin score histogram, min/max/mean spread, non-trivial differentiation between miners.
Validator & Scoring Robustness
Transparent scoring, strong resistance to manipulation/collusion, effective bias mitigation.
Evidenced by: critic quorum enforcement, EMA reputation factors, consensus-aligned critic identification, validator majority confirmation.
Rounds
Expand a round to see its full evidence trail