Human-Instruction-Free LLM Self-Alignment with Limited Samples

Open in new window