{"slug": "gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning", "title": "GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning", "summary": "Researchers introduced GEPA (Genetic-Pareto), a prompt optimizer that uses natural language reflection to improve LLM performance through trial and error. Across six tasks, GEPA outperformed reinforcement learning methods like GRPO by 6% on average and up to 20%, while using up to 35x fewer rollouts. The approach also surpassed the leading prompt optimizer MIPROv2 by over 10% and showed promise for code optimization.", "body_md": "# Computer Science > Computation and Language\n\n[Submitted on 25 Jul 2025 (\n\n[v1](https://arxiv.org/abs/2507.19457v1)), last revised 14 Feb 2026 (this version, v2)]# Title:GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning\n\n[View PDF](/pdf/2507.19457)\n\nAbstract:Large language models (LLMs) are increasingly adapted to downstream tasks via reinforcement learning (RL) methods like Group Relative Policy Optimization (GRPO), which often require thousands of rollouts to learn new tasks. We argue that the interpretable nature of language often provides a much richer learning medium for LLMs, compared to policy gradients derived from sparse, scalar rewards. To test this, we introduce GEPA (Genetic-Pareto), a prompt optimizer that thoroughly incorporates natural language reflection to learn high-level rules from trial and error. Given any AI system containing one or more LLM prompts, GEPA samples trajectories (e.g., reasoning, tool calls, and tool outputs) and reflects on them in natural language to diagnose problems, propose and test prompt updates, and combine complementary lessons from the Pareto frontier of its own attempts. As a result of GEPA's design, it can often turn even just a few rollouts into a large quality gain. Across six tasks, GEPA outperforms GRPO by 6% on average and by up to 20%, while using up to 35x fewer rollouts. GEPA also outperforms the leading prompt optimizer, MIPROv2, by over 10% (e.g., +12% accuracy on AIME-2025), and demonstrates promising results as an inference-time search strategy for code optimization. We release our code at[this https URL].\n\n## Submission history\n\nFrom: Lakshya A Agrawal [[view email](/show-email/af5f91b3/2507.19457)]\n\n**Fri, 25 Jul 2025 17:42:32 UTC (1,632 KB)**\n\n[[v1]](/abs/2507.19457v1)**[v2]** Sat, 14 Feb 2026 11:42:30 UTC (1,650 KB)\n\n### Current browse context:\n\ncs.CL\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer\n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers\n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps\n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations\n\n*(*[What are Smart Citations?](https://www.scite.ai/))# Code, Data and Media Associated with this Article\n\nalphaXiv\n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers\n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub\n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub\n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face\n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast\n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower\n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender\n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [ Learn more about arXivLabs](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning", "canonical_source": "https://arxiv.org/abs/2507.19457", "published_at": "2026-07-14 04:13:02+00:00", "updated_at": "2026-07-14 04:49:18.561347+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "ai-tools"], "entities": ["GEPA", "GRPO", "MIPROv2", "AIME-2025"], "alternates": {"html": "https://wpnews.pro/news/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning", "markdown": "https://wpnews.pro/news/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.md", "text": "https://wpnews.pro/news/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.txt", "jsonld": "https://wpnews.pro/news/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.jsonld"}}