FrogNano：通过在线任务合成训练 4B 编程智能体
FrogNano: Training a 4B Coding Agent via Online Task Synthesis
FrogNano 是一个 4B 编程智能体，仅通过 RL 在约 1500 个 SWE 环境的合成任务上后训练，无需从大模型蒸馏即可胜任软件工程任务。其核心是在线任务合成流水线，按当前 checkpoint 的学习前沿校准任务难度。该报告称在可学习前沿生成任务对性能提升很重要，目标是可在极低硬件上运行的轻量编程智能体。
Authors:Minseon Kim, Zhengyan Shi, Emiliano Penaloza, Christopher Cui, Roger Creus Castanyer, Maryam Hashemzadeh, Isadora White, Jonathan Light, Jeonghye Kim, Matheus Pereira, Darya Moldavskaya, Chinmay Singh, Fabio Vera, Baolin Peng, Xingdi Yuan, Marc-Alexandre Côté, Alessandro Sordoni
Abstract:We present FrogNano, a 4B coding agent designed to tackle software engineering (SWE) tasks efficiently and effectively, even under resource-constrained environments. It is post-trained exclusively via RL on around 1,500 SWE environments with synthetic tasks. A key ingredient for improving performance is an online task synthesis pipeline that creates tasks calibrated to the frontier of learnability for the current checkpoint. This report provides evidence that competitive small coding agents can be trained with synthetic tasks alone, without traditional distillation from larger models, and that generating tasks at the learnability frontier of the current agent is important. We report details on the training methodology, evaluations across diverse environments, and in-depth analyses, serving as a foundation for our ongoing exploration of lightweight yet capable coding agents that can run on minimal hardware.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.07925 [cs.AI] |
| (or arXiv:2609.07925v5 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2609.07925 arXiv-issued DOI via DataCite |
Submission history
From: Minseon Kim [view email]
[v1]
Mon, 7 Sep 2026 19:39:38 UTC (2,233 KB)
[v2]
Wed, 9 Sep 2026 15:35:02 UTC (2,216 KB)
[v3]
Mon, 14 Sep 2026 16:12:24 UTC (2,216 KB)
[v4]
Wed, 16 Sep 2026 17:38:11 UTC (2,219 KB)
[v5]
Mon, 5 Oct 2026 18:00:17 UTC (2,219 KB)
来源：arXiv：cs.AI · arxiv.org