StablePT: Towards Stable Prompting for Few-shot Learning via Input Separation

Open in new window