Boost, Disentangle, and Customize: A Robust System2-to-System1 Pipeline for Code Generation
Kounianhua Du, Hanjing Wang, Jianxing Liu, Jizheng Chen, Xinyi Dai, Yasheng Wang, Ruiming Tang, Yong Yu
Abstract
To address these limitations, we propose BDC, a novel framework that Boosts reasoning exploration via multi-agent collaboration, Disentangles heterogeneous data into specialized experts, and Customizes solutions through dynamic model composition. BDC integrates a Monte Carlo Tree-of-Agents algorithm, where multiple LLMs mutually verify and refine reasoning paths through reflection-guided pruning, enabling efficient exploration of high-quality solutions. To handle data diversity, we cluster problems by latent semantics, train composable LoRA experts on each cluster, and deploy an input-aware hypernetwork to dynamically merge these experts into tailored solvers. Experiments on APPS and CodeContest benchmarks demonstrate BDC’s superiority: it achieves up to 73.8% accuracy on hard problems, outperforming state-of-the-art methods like LATS and RethinkMCTS by 9–15%. This work lays the groundwork for advancing LLM capabilities in complex reasoning tasks, offering a novel System2-to-System1 solution.
BibTeX
@inproceedings{du-etal-2025-boost,
title = "Boost, Disentangle, and Customize: A Robust System2-to-System1 Pipeline for Code Generation",
author = "Du, Kounianhua and
Wang, Hanjing and
Liu, Jianxing and
Chen, Jizheng and
Dai, Xinyi and
Wang, Yasheng and
Tang, Ruiming and
Yu, Yong and
Wang, Jun and
Zhang, Weinan",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.833/",
doi = "10.18653/v1/2025.findings-acl.833",
pages = "16194--16204",
ISBN = "979-8-89176-256-5"
}