TrainExplorer: Autonomous LLM Post-Training via Plan-and-Decouple Exploration
LLM-based agents can automate model post-training, accelerating progress toward recursive self-improvement. Despite their promise, existing approaches remain under-explored and face two main challenges: (1) strategy lock-in that limits exploration beyond the initial strategy; and (2) coupled execution that obscures the effects of individual modifications. To address these challanges, we propose TrainExplorer, a multi-agent framework for autonomous post-training, that incorporates strategy planning and execution decoupling, to achieve broaden strategy exploration. First, strategy planning draws on scientific literature to construct a candidate search space of training strategies. Then, decoupled execution enables single-component interventions and attribution to guide strategy exploration. Extensive experiments on PostTrainBench demonstrate that TrainExplorer achieves a weighted aggregate score of 39.02, outperforming Claude Code using the same LLM by 7.32 and surpassing GPT-5.6 Sol-powered Codex, which scores 36.23. Code will be available.
Authors
- Zhengfeng Yu
- Feihu Jiang
- Qianghuai Jia (ORCID: https://orcid.org/0009-0009-6143-9950)
- Xuannan Liu (ORCID: https://orcid.org/0009-0005-1428-3261)
- Ran He (ORCID: https://orcid.org/0000-0002-3807-991X)
- Xing Cui
- Weihua Luo
- Longyue Wang
- Zhao Xu
- Tian Lan
- Peipei Li
- Zekun Li
Publication Details
- Journal
- arXiv (Cornell University)
- Published
- 2026-11-16
- DOI
- https://doi.org/10.48550/arxiv.2611.20430
- Primary Topic
- Software Engineering Research
- Type
- preprint