Harness-Aware Self-Evolving: Co-Evolving Model Weights, Harness, and Task Solutions
Abstract
Self-evolving frameworks usually optimize task solutions while treating the surrounding harness as fixed. We introduce Harness-Aware Self-Evolving (HASE), an agentic reinforcement-learning framework in which a single model can generate task solutions or edit selected harness components in a multi-turn action space. HASE enables a single Qwen3-8B model to match the text-classification performance of a GPT-OSS-120B model that uses Claude Code as the harness proposer. In alpha factor mining, HASE outperforms the reported GPT-OSS-120B baseline. HASE also repairs imperfect evaluation components and converges to state-of-the-art performance in circle-packing algorithm discovery. These results show that HASE improves the harness and the solution through one unified agentic process.
Cite
@article{arxiv.2607.03935,
title = {Harness-Aware Self-Evolving: Co-Evolving Model Weights, Harness, and Task Solutions},
author = {Haochen Luo and Yi Huang and Sichun Luo and Fengyuan Liu and Lei Li and Zefa Hu and Junlan Feng and Qi Liu},
journal= {arXiv preprint arXiv:2607.03935},
year = {2026}
}