While the ``deep reasoning'' paradigm has spurred significant advances in verifiable domains like mathematics, its application to open-ended, creative generation remains a critical challenge. The two dominant methods for instilling reasoning -- reinforcement learning (RL) and instruction distillation -- falter in this area; RL struggles with the absence of clear reward signals and high-quality reward models, while distillation is prohibitively expensive and capped by the teacher model's capabilities. To overcome these limitations, we introduce REverse-Engineered Reasoning (REER), a new paradigm that fundamentally shifts the approach. Instead of building a reasoning process ``forwards'' through trial-and-error or imitation, REER works ``backwards'' from known-good solutions to computationally discover the latent, step-by-step deep reasoning process that could have produced them. Using this scalable, gradient-free approach, we curate and open-source DeepWriting-20K, a large-scale dataset of 20,000 deep reasoning trajectories for open-ended tasks. Our model, DeepWriter-8B, trained on this data, not only surpasses strong open-source baselines but also achieves performance competitive with, and at times superior to, leading proprietary models like GPT-4o and Claude 3.5.
@article{arxiv.2509.06160,
title = {Reverse-Engineered Reasoning for Open-Ended Generation},
author = {Haozhe Wang and Haoran Que and Qixin Xu and Minghao Liu and Wangchunshu Zhou and Jiazhan Feng and Wanjun Zhong and Wei Ye and Tong Yang and Wenhao Huang and Ge Zhang and Fangzhen Lin},
journal= {arXiv preprint arXiv:2509.06160},
year = {2025}
}