Are Human-generated Demonstrations Necessary for In-context Learning?

Open in new window