simple-mathconditionalweight 0.5adapter 7 of 17SimpleMath Evals.
The SimpleMath adapter runs the SimpleMath baseline evaluation: short arithmetic problems with deterministic answers. It is a floor check rather than a depth test — it verifies the subject can perform exact reasoning at the simplest level, and is weighted accordingly.
What it measures.
Problem correctness
Share of arithmetic problems answered exactly right.
Deterministic scoring
Answers are checked exactly, no judge model involved.
Component keys.
simple-math.scoreNormalized correctness score on the 0–100 component scale.
When it counts.
Conditional — counts when the subject has SimpleMath evaluation results.
When signals are missing.
No eval results means the component is missing; the composite uses the adapters that have signal.
How it contributes.
{
"component": "simple-math.*",
"scale": "0-100",
"mode": "conditional",
"weight": 0.5,
"contribution": "component x 0.5 / total weight"
}The adapter’s components are averaged into the composite as a weighted mean: every contributing adapter’s component is multiplied by its weight, summed, and divided by the total weight of contributing adapters. The result is published as trustScores.total with a config version and timestamp so anyone can recompute it.
Questions about SimpleMath Evals.
What is the SimpleMath adapter?
It converts SimpleMath evaluation results — exact-answer arithmetic problems — into the simple-math.score component on the 0–100 scale.
Why is SimpleMath weighted at 0.5?
Arithmetic correctness is a useful floor check but says little about real-world capability, so it carries a reduced default weight in the informative catalog.