harnyGlossary
Procedural memory
Procedural memory is an AI agent's memory of how to do something: the steps, rules and routines it learned to follow for a task. It sits apart from memory of facts and past events.
What procedural memory is
Procedural memory stores know-how. An agent uses it to recall the order of steps for a task, the rules that apply, and the way a job was done before.
It is the agent equivalent of knowing how to ride a bike, not just remembering that you once did.
Why it matters
- It lets an agent repeat good routines without being walked through them again.
- It is hard to get right, because a stored procedure can go stale when the steps or rules change.
- It needs care about which rules an agent may follow on its own.
How it works in Harny
- Harny's learned settings are the nearest thing. When recorded responses support it, Harny offers a change to an action kind's setting. You apply it or keep the current one, and you can restore the previous setting.
- The source is recorded responses, not a trained model.
Is procedural memory the same as a learned setting in Harny?
No. A learned setting is an offered change to an action kind's setting, based on recorded responses. You apply it or keep the current one.
Keep reading
Go on. Get Harny.
The first 500 monthly-active end users are free for good. Create an app key and make your first call in minutes.