3 ms·DeepMind discovers that AI large language models can optimize their own prompts26 points by labrador 3y agofalcor84 3y agoThe original paper on arxiv: https://arxiv.org/abs/2309.03409 https://arxiv.org/abs/2309.03409