# ZERA: Zero-init Instruction Evolving Refinement Agent -- From Zero Instructions to Structured Prompts via Principle-based Optimization

arXiv: https://arxiv.org/abs/2509.18158 · HTML: https://mlsanity.com/paper/2509.18158/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use.
- Builds on 17 of its 23 references that are in the map.
- First posted on arXiv 17 Sep 2025.

## Placed in

- LLM agents and tool use / [General LLM agent frameworks and benchmarks](https://mlsanity.com/agents/general-llm-agent-frameworks-and-benchmarks/)

## Builds on (most cited first, 17 of 17)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them](https://mlsanity.com/paper/2210.09261/) (2022)
- [Calibrate Before Use: Improving Few-Shot Performance of Language Models](https://mlsanity.com/paper/2102.09690/) (2021)
- [Teaching Machines to Read and Comprehend](https://mlsanity.com/paper/1506.03340/) (2015)
- [Large Language Models Are Human-Level Prompt Engineers](https://mlsanity.com/paper/2211.01910/) (2022)
- [Automatic Prompt Optimization with "Gradient Descent" and Beam Search](https://mlsanity.com/paper/2305.03495/) (2023)
- [True Few-Shot Learning with Language Models](https://mlsanity.com/paper/2105.11447/) (2021)
- [PromptAgent: Strategic Planning with Language Models Enables Expert-level Prompt Optimization](https://mlsanity.com/paper/2310.16427/) (2023)
- [Self-Supervised Prompt Optimization](https://mlsanity.com/paper/2502.06855/) (2025)
- [MAPO: Boosting Large Language Model Performance with Model-Adaptive Prompt Optimization](https://mlsanity.com/paper/2407.04118/) (2024)
- [MORL-Prompt: An Empirical Analysis of Multi-Objective Reinforcement Learning for Discrete Prompt Optimization](https://mlsanity.com/paper/2402.11711/) (2024)
- [Revisiting Prompt Optimization with Large Reasoning Models-A Case Study on Event Extraction](https://mlsanity.com/paper/2504.07357/) (2025)
- [Putting People in LLMs' Shoes: Generating Better Answers via Question Rewriter](https://mlsanity.com/paper/2408.10573/) (2024)

## Built on by (most cited first, 6 of 6)

- [Adaptive Prompt Structure Factorization: A Framework for Self-Discovering and Optimizing Compositional Prompt Programs](https://mlsanity.com/paper/2604.06699/) (2026)
- [Prompt Optimization Enables Stable Algorithmic Collusion in LLM Agents](https://mlsanity.com/paper/2604.17774/) (2026)
- [PIAST: Rapid Prompting with In-context Augmentation for Scarce Training data](https://mlsanity.com/paper/2512.11013/) (2025)
- [Communication Policy Evolution for Proactive LLM Agents](https://mlsanity.com/paper/2606.14314/) (2026)
- [MIDAS: Multi-LLM Iterative Data-Adaptive Summarization](https://mlsanity.com/paper/2608.04307/) (2026)
- [SEPO: Evidence-Grounded Prompt Optimization via Structural Editing](https://mlsanity.com/paper/2608.28067/) (2026)

## Abstract (arXiv)

Automatic Prompt Optimization (APO) improves large language model (LLM) performance by refining prompts for specific tasks. However, prior APO methods typically focus only on user prompts, rely on unstructured feedback, and require large sample sizes and long iteration cycles-making them costly and brittle. We propose ZERA (Zero-init Instruction Evolving Refinement Agent), a novel framework that jointly optimizes both system and user prompts through principled, low-overhead refinement. ZERA scores prompts using eight generalizable criteria with automatically inferred weights, and revises prompts based on these structured critiques. This enables fast convergence to high-quality prompts using minimal examples and short iteration cycles. We evaluate ZERA across five LLMs and nine diverse datasets spanning reasoning, summarization, and code generation tasks. Experimental results demonstrate consistent improvements over strong baselines. Further ablation studies highlight the contribution of each component to more effective prompt construction. Our implementation including all prompts is publicly available at https://github.com/younatics/zera-agent.