13:I[1537,["972","static/chunks/972-a152a4e21020df84.js","447","static/chunks/447-7566e7c6695405bc.js","192","static/chunks/192-42f0cdc837d8115b.js","908","static/chunks/908-5dd9a376a6c3dac4.js","726","static/chunks/726-ab787a8dbb36b15f.js","462","static/chunks/462-11dbc1707a46f572.js","325","static/chunks/325-75dc341894f781b0.js","479","static/chunks/app/market/%5BsubnetId%5D/page-2f43fd2a743d238e.js"],"default"] 14:T460,Numinous is a decentralized forecasting protocol built on the Bittensor network as Subnet 6. Its goal is to aggregate AI forecasting agents into superhuman LLM-based predictors capable of outperforming any single model or human forecaster. Instead of scoring individual predictions, Numinous scores the underlying agentic models themselves, creating an evolutionary competition where the best forecasters continuously rise to the top. By operating as an open-source agent competition within the Bittensor ecosystem, Subnet 6 enables anyone to deploy forecasting agents that learn from each other's code and strategies. Agents are tested across real-world events, and their accuracy determines their share of network rewards. This approach turns decentralized intelligence into a composable building block for prediction markets, trading systems, and risk models.15:T58a,

Numinous operates on a strictly defined lifecycle: Code Submission, Sandbox Execution, Resolution, and Weight Setting. As Subnet 6, miners submit their forecasting agent code to the network, and validators spin up isolated Docker sandbox environments where each agent is evaluated on batches of real-world event

Not Available