Proof of Alpha at a Glance

Contribute the most useful alpha 
to gain influence and rewards.

Proof of Alpha is the core reward mechanism powering the Upshot Machine Intelligence Network. It ensures network participants are rewarded based on how much useful alpha they contribute.

Upshot MIN participants contribute alpha in the form of proprietary data, predictive features, machine learning models, and other elements aimed at enhancing the network’s ability to optimize specific objective functions. The participants who contribute the most useful alpha will earn the most rewards.

Proof of Alpha works through the following steps:

  1. Alpha Miners stake (or receive delegated stakes from others) backing their models, indicating confidence in their alpha. More stake means higher potential rewards or losses.
  2. Network participants score other models in the network based on how well they perform relative to that subnetwork’s underlying objective functions (e.g. Mean Absolute Directional Loss or Max Sharpe Ratio).
  3. Rewards are generated from protocol fees and inflation, and alpha miner payouts get distributed proportional to each model's accuracy. More accurate models earn higher rewards, while inaccurate ones may be penalized. Additionally, participants are rewarded based on the scores they give to other models in the network.
  4. Models are then combined into a weighted, aggregated "meta-model" that produces consensus price predictions.

The result is a system akin to a perpetual prediction market, with regular rewards distributed based on contribution quality and accuracy. Proof of Alpha ensures the most effective models for each subnetwork are compensated accordingly.