Question 1

What is Crene?

Accepted Answer

Crene is probability infrastructure for institutional decisions. The platform produces calibrated probability data on forward looking events where pricing mechanisms do not exist, including macro releases, earnings, central bank decisions, commodities, crypto, and policy events. Each event is forecast independently by four frontier AI models (Claude, GPT, Gemini, Grok); cross model disagreement is exposed as a measured uncertainty signal. Every resolved event is closed against a tier A authoritative source.

Question 2

How does AI prediction calibration work at Crene?

Accepted Answer

Calibration applies at the consensus layer, not at the underlying model layer. Crene queries four frontier models in isolation, computes a cross model median consensus, and tracks resolution against a tiered source allowlist. Every resolved event becomes a permanent calibration point. Per tier accuracy, per model Brier scores, and per domain accuracy are computed live and published at https://crene.com/methodology.

Question 3

What is a Brier score?

Accepted Answer

A Brier score is the proper scoring rule for probability forecasts on binary outcomes. It is the mean squared error between the forecast probability and the realized outcome (0 or 1). Lower scores indicate better calibration. The no skill baseline is 0.25 (always predicting 0.5); a perfect forecaster scores 0.0. Crene publishes the consensus Brier score and per model Brier scores across the resolved event corpus.

Question 4

Who is Crene for?

Accepted Answer

Crene serves institutional users: research desks, allocators, risk teams, AI agents that require probabilistic reasoning, and internal decision systems that need credible uncertainty quantification. Dataset access is available three ways: direct license (resolved event archive, calibration history, cluster decompositions), data marketplaces (Neudata, Eagle Alpha, Monda), and the public read API at https://api-get.crene.com. The Probability endpoint for arbitrary forward looking questions is in private design partner development.

Question 5

What is the difference between an event, a cluster, and a factor at Crene?

Accepted Answer

Events, clusters, and factors are the three primitives. An event is a base binary outcome with a scalar probability between 0 and 1. A cluster is a composition of related events, a thesis decomposed into its binary primitives with consensus and disagreement tracked at both levels. A factor is the continuous sibling of an event: a state variable forecast as a cross model percentile distribution (p5, p25, p50, p75, p95) rather than a scalar, with a driver decomposition that explains movement in the parent distribution without generating it.

Will Hugging Face report more than 25% of enterprise inference workloads running open-source models in 2026?

Belief trajectory

Per-model probabilities

Resolution criteria

Will Hugging Face report more than 25% of enterprise inference workloads running open-source models in 2026?

Belief trajectory

Per-model probabilities

Resolution criteria