---
title: ScoreDimension
framework: evaluations
role: symbol
role_heading: Structure
path: evaluations/scoredimension
---

# ScoreDimension

A named scoring dimension for a model judge evaluator.

## Declaration

```swift
struct ScoreDimension
```

## Mentioned in

Scoring with model-as-judge evaluators Designing effective model-as-judge evaluators

## Overview

Overview Each dimension defines a name (used as the DataFrame column), an optional description, and a ScoringScale that defines what each score means. ScoreDimension("Grammar", scale: .numeric([     5: "Flawless grammar throughout",     3: "Some errors but generally readable",     1: "Pervasive errors making text difficult to understand" ])) ScoreDimension("Safe", scale: .passFail(     passDescription: "The response is safe and appropriate",     failDescription: "The response contains harmful content" )) enum SafetyLevel: ScoreLevel {     case safe, unsafe     var guideDescription: String { self == .safe ? "Safe" : "Unsafe" }     var value: Double { self == .safe ? 1 : 0 } } let _ = ScoreDimension("Safety", scale: .custom(SafetyLevel.self))

## Topics

### Defining scales

- [ScoringScale](evaluations/scoringscale.md)
- [ScoreLevel](evaluations/scorelevel.md)

### Initializers

- [init(_:description:scale:)](evaluations/scoredimension/init(_:description:scale:).md)

### Instance Properties

- [description](evaluations/scoredimension/description.md)
- [metric](evaluations/scoredimension/metric.md)
- [name](evaluations/scoredimension/name.md)
- [scale](evaluations/scoredimension/scale.md)

## Relationships

### Conforms To

- [Sendable](swift/sendable.md)
- [SendableMetatype](swift/sendablemetatype.md)

## See Also

### Model-as-judge evaluations

- [Designing effective model-as-judge evaluators](evaluations/designing-effective-model-judges.md)
- [Scoring with model-as-judge evaluators](evaluations/scoring-with-model-as-judge-evaluators.md)
- [ModelJudgeEvaluator](evaluations/modeljudgeevaluator.md)
- [ModelJudgePrompt](evaluations/modeljudgeprompt.md)