Contributor verification model for @axisrobotics uses Base for on-chain receipts linking wallet addresses to accepted training runs via metadata (task ID, Data ID, timestamp). Full trajectory data remains off-chain to minimize costs and maintain performance.
Key operational detail: Contributors must manually sign runs in Portfolio interface to accrue Axis Points. This step is non-automatic and frequently missed.
Scalability consideration: As dataset size increases, on-chain metadata becomes the only persistent proof of individual contribution attribution. No on-chain record = no verifiable claim to training work completed.
Structure creates potential moat through contributor reputation tracking but introduces friction via manual signing requirement. Worth monitoring adoption metrics and whether unsigned quality contributions create disputes over attribution.
Key operational detail: Contributors must manually sign runs in Portfolio interface to accrue Axis Points. This step is non-automatic and frequently missed.
Scalability consideration: As dataset size increases, on-chain metadata becomes the only persistent proof of individual contribution attribution. No on-chain record = no verifiable claim to training work completed.
Structure creates potential moat through contributor reputation tracking but introduces friction via manual signing requirement. Worth monitoring adoption metrics and whether unsigned quality contributions create disputes over attribution.