> ## Documentation Index
> Fetch the complete documentation index at: https://logicfold.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Your Memory

> See and tune exactly what Logic AI remembers about you.

As you chat, Logic AI quietly remembers things that help it work better for you — your role, your preferences, terminology you use, the kinds of records you work with. It recalls them automatically in later conversations, so you don't have to re-explain yourself.

You're always in control of what it keeps. The **Memory** panel shows everything Logic AI has learned about you, and lets you confirm, correct, or forget any of it.

## Opening your memory

In the chat header, click **Memory** ("See & tune what I remember about you"). The panel opens in place of the chat; use the back arrow to return. Everything here is **private to you** — no one else sees your memory.

## The overview

At the top, two quick metrics:

* **Accuracy** — a score reflecting how well your memory is serving you, based on your feedback.
* **Reviewed** — how many of your remembered facts you've checked, with a progress bar. Reviewing helps Logic AI trust the right facts.

Below that you can **filter** by category using the pills, or **search** your memories by keyword.

## Reviewing what Logic AI remembers

Memories are shown as cards, grouped into **Needs review** (new facts Logic AI hasn't had you confirm yet) and **Reviewed** (ones you've already checked). Each card shows the fact and its category. On any card you can:

* **👍 / 👎 — "Is this right?"** Confirm a fact is correct, or flag that it's off. Your verdict tunes how much Logic AI trusts it.
* **✏️ Edit** — fix the wording of a fact so it's exactly right.
* **✕ Forget** — remove a fact you don't want kept. You'll be asked to confirm; once forgotten, Logic AI won't silently re-learn it.

When you've checked everything, you'll see **"You're all caught up."**

## Tuning memory as you chat

You don't have to open the panel to steer memory. When you 👍 / 👎 a message in the chat, Logic AI shows a few quick **memory feedback** chips ("What worked?" or "Was it my memory?") so you can tell it, in one click, whether its memory helped. See [Feedback & Reactions](/products/logicai/docs/feedback-and-reactions).

## When memory is off

If your admin has turned memory off for the org, the panel says so and Logic AI won't remember anything between conversations. An admin can enable it from the [Memory](/products/logicai/docs/admin-memory) tab.

## Coworker memory

[Coworkers](/products/logicai/docs/coworkers) have their own separate memory, reviewed the same way — open a coworker's memory from its menu in the sidebar. A shared coworker's memory is shared with everyone who can use it, so curating it is a team effort.