chanmuzi
<LK Lab, Retrieval> [RoSPr] Efficiently Enhancing Zero-Shot Performance of Instruction Following Model via Retrieval of Soft Prompt (2023.10)