NPO用简单方法实现高效提示优化，教师模型越强效果越好，颠覆了复杂优化器的设计思路。
NPO是一种单谱系提示优化器，在指令跟随基准上仅需3,500和6,800次轮次就达到GEPA的3,593和6,871次轮次的性能。NPO不使用候选种群和搜索树，而是通过滑动窗口将最近迭代交给教师模型重写提示。在22个TextArena游戏中，NPO保持与GEPA相当的性能。随着教师模型能力增强，NPO的优势会扩大，这表明优化器端的搜索复杂度一直在弥补教师推理能力的不足。
Interesting paper on prompt optimization. They claim that a single-lineage prompt optimizer just ma...
Interesting paper on prompt optimization. They claim that a single-lineage prompt optimizer just matched GEPA on a smaller rollout budget. Prompt optimization has been drifting toward heavier machinery, with candidate pools, reflection trees, and Pareto-based selection. NPO keeps one lineage. At each iteration it runs the student on the current prompt, collects rollout traces and rewards, and hands a sliding window of recent iterations to a teacher model that rewrites the prompt. There is no candidate population and no search tree. On the two instruction-following benchmarks it spends 3,500 and 6,800 rollouts against GEPA's 3,593 and 6,871, and it stays broadly comparable across 22 TextArena games. The interaction with teacher strength is what makes this interesting. NPO's advantage grows as the teacher model gets stronger, which suggests optimizer-side search complexity has been compensating for weak teacher reasoning all along. Paper: arxiv.org/abs/2608.27266 Chat with Paper: academy.dair.ai/papers/naive-p… 💬 1 🔄 3 ❤️ 20 👀 2144 📊 7 ⚡