In-context Learning
Pembelajaran Dalam Konteks
The model adapts to a task from the context alone, without any training.
In-context learning is the model's ability to pick up a task from what is in the window: instructions, examples, a format, a correction. No weights change. The adaptation happens per request and disappears when the request ends.
This is more capable than it first sounds, and it is the reason a model can follow a house style it has never been trained on, or fix its own mistake after you point it out mid-conversation.
It also has a limit worth knowing: the model does not accumulate skill across requests. It cannot learn your codebase by reading it once, which is why tools, memory, and retrieval exist. In-context learning is powerful, but it is also perishable.
Related terms
Few-shot Prompting
Showing the model two or three examples of the task instead of describing it.
Context Window
Everything the model sees on one request. Finite, and it decides what fits.
Memory
A system's ability to carry information from one request to the next.
Fine-tuning
Further training of an existing model on your own data to shape its behaviour.
LLM
A model trained on a huge amount of text to predict the next token. The engine behind most AI products.
Want this working inside your team?
It starts with one conversation about your workflow.