Modeling DeFi forward conviction
The GM! Index, modeling DeFi forward conviction with the DataFi likelihood-native foundation model. The DataFi Forward Conviction Index, premier GM! Index as featured index components. The DataFi Forward Conviction Index is the set of all instruments, horizons, granularities, and action levels. GM! Index is the subset whose likelihood distributions achieve the top forward information coefficient readings with confidence. Each cell scores 0.0 (max short) → 1.0 (max long) with confidence, plus an additive FCT omega triad (mass, direction, ΔΩ) and θ stamps. Production primary remains composite so live traders stay byte-identical.
The research program defines a validation tier that separates feed quality from profit-trader PnL, with objective profiles that pass the feed gate before publication.
A quick surf is a pay-per-read of the published GM! Index (omega book and omega market line). A session is a batch on the desk. COMPUTER is the desk. X is the public room. Public preview, then x402, then the house-banked desk. Book DFY later. The mint is not required. Token-gating is not required.
- Tier C feed construction is bearer-gated — public surfaces show conviction only
- Non-custodial data · not investment advice · USDC x402 or house-banked desk turns on Base · mint is not required
GM! Index Feed
Live public showcase · default oracle: BTC1D/DFY
Mint
00–38 of 4,440. Founder collection of 100 on Base. Public preview is a silhouette — mint unlocks the image. Read access is a refillable entitlement, not a lock on published data.
API & integration
HTTP, MCP, x402, ACP, entitlement metadata, and connector reference for builders and agents

