What Affine (Subnet 120) Does in Bittensor

Interesting article from Gab about Affine (SN120) and how it’s positioning as the reinforcement learning (RL) engine for Bittensor.

Affine (Subnet 120) in the Bittensor ecosystem uses reinforcement learning (RL) to enable miners to compete in improving AI models for tasks like reasoning and coding.

It incorporates cheat-proof mechanisms, such as resistance to Sybil attacks and overfitting, to ensure fair competition.

Through integration with Subnet 64 (Chutes), it facilitates open model deployment, promoting decentralized AI advancement and challenging centralized dominance in the field.

Below is an image breaking down what Affine does:

Click here to read the full article.