---
title: "Labeling correct outcomes"
description: "How can you define correctness criteria instead of judging overall impressions?"
status: "coming-soon"
language: "en"
canonical: "https://bulnik.dev/en/articles/topics/T58.02/"
topicId: "T58.02"
---

> Coming soon — this article is being prepared. Below are its question, intended outcome, and place in the story.

## The question

How can you define correctness criteria instead of judging overall impressions?

## What we will work through

Define correctness criteria instead of judging overall impressions.

## Before you begin

Previously covered: Datasets for evaluating an AI feature; Facts, interpretations, and opinions.

- [Datasets for evaluating an AI feature](/en/articles/topics/T58.01/)
- [Facts, interpretations, and opinions](/en/articles/topics/N31.01/)