The open-source AI research organization Nous Research announced the establishment of a subnet on Bittensor

ChainCatcher news, the open-source AI research organization Nous Research announced on social media the establishment of a subnet on the decentralized AI network project Bittensor (TAO), aiming to provide a new evaluation system for open-source models.

Nous Research stated that traditional benchmarking heavily relies on public datasets, which are easily exploited and often lead to scores that appear high but obscure the true capabilities of the models, misleading developers into meaningless score chasing.

It is reported that the Cortex subnet is a dynamic source of fresh synthetic data continuously generated by GPT-4, allowing the use of fresh and unpredictable high-quality data to test models and ensuring a fair and accurate consensus on which model most closely reflects GPT-4's performance. This new system aims to celebrate and reward the creators of truly user-needs-satisfying public open-source models through Bittensor's built-in incentive structure.