# System Prompt Optimization with Meta-Learning

arXiv: https://arxiv.org/abs/2505.09666 · HTML: https://mlsanity.com/paper/2505.09666/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Builds on 35 of its 46 references that are in the map.
- First posted on arXiv 14 May 2025.

## Builds on (most cited first, 30 of 35)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://mlsanity.com/paper/1703.03400/) (2017)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [The Power of Scale for Parameter-Efficient Prompt Tuning](https://mlsanity.com/paper/2104.08691/) (2021)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [Matching Networks for One Shot Learning](https://mlsanity.com/paper/1606.04080/) (2016)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [Large Language Models Are Human-Level Prompt Engineers](https://mlsanity.com/paper/2211.01910/) (2022)
- [MPNet: Masked and Permuted Pre-training for Language Understanding](https://mlsanity.com/paper/2004.09297/) (2020)
- [DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs](https://mlsanity.com/paper/1903.00161/) (2019)
- [Large Language Models as Optimizers](https://mlsanity.com/paper/2309.03409/) (2023)
- [Automatic Prompt Optimization with "Gradient Descent" and Beam Search](https://mlsanity.com/paper/2305.03495/) (2023)
- [Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution](https://mlsanity.com/paper/2309.16797/) (2023)
- [MedMCQA : A Large-scale Multi-Subject Multi-Choice Dataset for Medical domain Question Answering](https://mlsanity.com/paper/2203.14371/) (2022)
- [The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions](https://mlsanity.com/paper/2404.13208/) (2024)
- [TweetEval: Unified Benchmark and Comparative Evaluation for Tweet Classification](https://mlsanity.com/paper/2010.12421/) (2020)
- ["Liar, Liar Pants on Fire": A New Benchmark Dataset for Fake News Detection](https://mlsanity.com/paper/1705.00648/) (2017)
- [PromptAgent: Strategic Planning with Language Models Enables Expert-level Prompt Optimization](https://mlsanity.com/paper/2310.16427/) (2023)
- [HateCheck: Functional Tests for Hate Speech Detection Models](https://mlsanity.com/paper/2012.15606/) (2020)
- [Multitask Prompt Tuning Enables Parameter-Efficient Transfer Learning](https://mlsanity.com/paper/2303.02861/) (2023)
- [Aligning to Thousands of Preferences via System Message Generalization](https://mlsanity.com/paper/2405.17977/) (2024)
- [ETHOS: an Online Hate Speech Detection Dataset](https://mlsanity.com/paper/2006.08328/) (2020)
- [When "A Helpful Assistant" Is Not Really Helpful: Personas in System Prompts Do Not Improve Performances of Large Language Models](https://mlsanity.com/paper/2311.10054/) (2023)
- [PRompt Optimization in Multi-Step Tasks (PROMST): Integrating Human Feedback and Heuristic-based Sampling](https://mlsanity.com/paper/2402.08702/) (2024)

## Built on by (most cited first, 18 of 18)

- [On-Policy Context Distillation for Language Models](https://mlsanity.com/paper/2602.12275/) (2026)
- [Prompt-MII: Meta-Learning Instruction Induction for LLMs](https://mlsanity.com/paper/2510.16932/) (2025)
- [Multimodal Prompt Optimization: Why Not Leverage Multiple Modalities for MLLMs](https://mlsanity.com/paper/2510.09201/) (2025)
- [MEMO: Memory-Augmented Model Context Optimization for Robust Multi-Turn Multi-Agent LLM Games](https://mlsanity.com/paper/2603.09022/) (2026)
- [When Thoughts Meet Facts: Reusable Reasoning for Long-Context LMs](https://mlsanity.com/paper/2510.07499/) (2025)
- [SePO: Self-Evolving Prompt Agent for System Prompt Optimization](https://mlsanity.com/paper/2606.04465/) (2026)
- [When Context Returns: Toward Robust Internalization in On-Policy Distillation](https://mlsanity.com/paper/2606.11627/) (2026)
- [Efficient Prompt Optimisation for Legal Text Classification with Proxy Prompt Evaluator](https://mlsanity.com/paper/2510.08524/) (2025)
- [Adaptive Multi-Agent Reasoning via Automated Workflow Generation](https://mlsanity.com/paper/2507.14393/) (2025)
- [Beyond Elicitation: Provision-based Prompt Optimization for Knowledge-Intensive Tasks](https://mlsanity.com/paper/2511.10465/) (2025)
- [Automated Reformulation of Robust Optimization via Memory-Augmented Large Language Models](https://mlsanity.com/paper/2605.11813/) (2026)
- [Training Language Agents to Learn from Experience](https://mlsanity.com/paper/2605.20477/) (2026)
- [EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents](https://mlsanity.com/paper/2609.17632/) (2026)
- [Enhancing Factual Accuracy and Citation Generation in LLMs via Multi-Stage Self-Verification](https://mlsanity.com/paper/2509.05741/) (2025)
- [LegalMidm: Use-Case-Driven Legal Domain Specialization for Korean Large Language Model](https://mlsanity.com/paper/2604.25297/) (2026)
- [VTOS: Learning to Orchestrate Vision Tools by Co-Searching Solutions and Observers](https://mlsanity.com/paper/2606.20728/) (2026)
- [Fair on the Surface? Benchmarking Hidden-Output Fairness Gaps in LLM Recommenders](https://mlsanity.com/paper/2608.08284/) (2026)
- [One Prompt Does Not Fit All: Self-Meta-Evolve for Personalized Information Extraction](https://mlsanity.com/paper/2609.21626/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) have shown remarkable capabilities, with optimizing their input prompts playing a pivotal role in maximizing their performance. However, while LLM prompts consist of both the task-agnostic system prompts and task-specific user prompts, existing work on prompt optimization has focused on user prompts specific to individual queries or tasks, and largely overlooked the system prompt that is, once optimized, applicable across different tasks and domains. Motivated by this, we introduce the novel problem of bilevel system prompt optimization, whose objective is to design system prompts that are robust to diverse user prompts and transferable to unseen tasks. To tackle this problem, we then propose a meta-learning framework, which meta-learns the system prompt by optimizing it over various user prompts across multiple datasets, while simultaneously updating the user prompts in an iterative manner to ensure synergy between them. We conduct experiments on 14 unseen datasets spanning 5 different domains, on which we show that our approach produces system prompts that generalize effectively to diverse user prompts. Also, our findings reveal that the optimized system prompt enables rapid adaptation even to unseen tasks, requiring fewer optimization steps for test-time user prompts while achieving improved performance.