# What is Working Memory?

It is the short-term memory area where artificial intelligence temporarily keeps information during processing.

## Overview
Working memory is where the AI ​​keeps the current context and necessary data active while performing a task. This information is usually cleared when the task is finished or the context changes.

*Analogy: It's like temporary notes you make on the side of a paper while solving a math problem; When the problem is over, those notes won't matter.*

## How it works
It occurs when the model keeps input data and intermediate results in its memory for quick access within the current processing cycle.

## Where it is used
It is used in preserving chat history and complex reasoning processes.

## Commonly confused with
Its difference from long-term memory is that it is limited only to the current task.

## Frequently asked questions
**What happens if working memory becomes full?**
The model begins to forget old information or may lose context.


## Related terms
- [Memory](/en/dictionary/memory/)
- [Context Window](/en/dictionary/context-window/)
- [Long-term Memory](/en/dictionary/long-term-memory/)

---
Source: TreScout Dictionary · https://trescout.com/en/dictionary/working-memory/