1 result for “in-context learning”
In-context learning is the ability of large language models to perform new tasks by conditioning on examples or instructions provided within the input prompt, without updating model weights.