跳到正文
arXiv cs.AI· Jiaxin Zhang, Yao Luo, Yuze Dan, Yunqin Li, Bowen Wang·· 56 分钟前AI 评分29

IfcGym:用可验证检查器作为训练信号,训练依赖感知的 IFC 建筑模型编辑智能体

Learning to Edit IFC Building Models: Verifiable Checkers as the Training Signal for Dependency-Aware Editing Agents

AI 导读

研究者提出 IfcGym,一个让大语言模型通过可执行代码编辑 IFC 建筑模型的 CodeAct 环境,用四个检查器验证有效性、意图、依赖与保留约束,其合取构成严格成功标准。

正文 · 原文

View PDF HTML (experimental)

Abstract:Editing a building information model in the Industry Foundation Classes (IFC) schema requires propagating changes to dependent elements while preserving unrelated entities. Existing language-model agents for IFC editing use prompted models without learning from edit outcomes. Target-file comparisons do not specify which alternative edits are allowed or which unrelated entities must remain unchanged. We present IfcGym, a CodeAct environment in which a language model edits IFC through executable code and four checkers verify validity relative to the input model, intent, dependency and preservation against a task specification. Their conjunction is strict success. IfcEdit distils verified teacher episodes into a Qwen3.5-9B student through supervised fine-tuning (SFT), then applies Group Relative Policy Optimization (GRPO) with strict success as the reward. On IfcEditBench, 1,990 editing tasks on 13 held-out real projects, the final student reaches 86.78% strict success, from 79.15% after SFT. The paired confidence interval for the difference from its teacher includes zero; the same recipe on a 27B model reaches 91.76% and exceeds the teacher. On the public benchmarks BIM-Edit and BIBIMBAP the final student scores 55.19 and 54.50. On IfcEditBench, the students outperform all prompted models except the teacher under matched tool access; both evaluated frameworks also score lower. Rewarding all four checkers yields 4.83 points more strict success than rewarding intent and validity alone.
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.22307 [cs.AI]
  (or arXiv:2611.22307v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.22307

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Bowen Wang [view email]
[v1] Tue, 17 Nov 2026 07:06:17 UTC (972 KB)

来源:arXiv cs.AI · arxiv.org