Accuracy & baselines

In plain words

How accurate are we? Here is a live score: how often the system’s forecasts come true overall, and for each AI on its own. The main point: the shared decision of several AIs is steadier than any one of them alone — that is why we use a team.

Click a percentage to open the full row list and browser recalculation.

Standard (refusals out of score)

HIGH_UNCERTAINTY is uncertainty (was: refusal) — not in accuracy. See the last column.

Model Accuracy % Evaluated total Correct Partial Wrong Uncertainty (was: Refusals)

Confusion matrix (all models)

Model rotation log

When Model Field Old New