Research map / Large language models / Agents, code and math
CME-CAD: Heterogeneous Collaborative Multi-Expert Reinforcement Learning for CAD Code Generation
Ke Niu, Haiyang Yu, Zhuofan Chen, Zhengtao Yao, Weitao Jia, Xiaodong Ge, Jingqun Tang, Benlei Cui, Bin Li, Xiangyang Xue
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train on multi-expert CoT data in SFT instead of a single expert (ablation) | Qwen3-VL-4B · CADExpert | IoU (%): 40.91 -> 64.45 (Expert3) | Improved |
| add GRPO after single-expert SFT (ablation) | Qwen3-VL-4B single expert · CADExpert | IoU (%): 40.91 -> 52.38 (Expert3) | Improved |
| add multi-expert RL (full CME-CAD) over single-expert SFT+GRPO (ablation) | Qwen3-VL-4B · CADExpert | IoU (%): 52.38 -> 80.71 (Expert3) | Improved |
| add Expert-Internal Advantage Estimation (ablation) | CME-CAD · CADExpert | IoU (%): 64.45 -> 73.89 | Improved |
| add Expert-Internal Advantage Estimation (ablation) | CME-CAD · CADExpert | Executability (%): 96.99 -> 98.28 | Improved |
| add hard negative sample buffering on top of EIAE (ablation) | CME-CAD · CADExpert | IoU (%): 73.89 -> 78.14 | Improved |
| add multi-expert collaborative learning on top of EIAE (ablation) | CME-CAD · CADExpert | IoU (%): 73.89 -> 76.71 | Improved |
| add multi-expert collaborative learning to EIAE+HSB (ablation) | CME-CAD · CADExpert | IoU (%): 78.14 -> 80.71 | Improved |
| add HSB or MECL (ablation) | CME-CAD · CADExpert | Executability (%): 98.28 -> 98.22 / 98.19 | No clear effect |
| use CME-CAD instead of CAD-RL | Qwen3-VL-4B · CADExpert | IoU (%): 71.84 -> 80.71 | Improved |
Closest papers by meaning
- From Intent to Execution: Multimodal Chain-of-Thought Reinforcement Learning for Precise CAD Code Generation (AAAI 2026 · similarity 0.93)
- cadrille: Multi-modal CAD Reconstruction with Reinforcement Learning (ICLR 2026 · similarity 0.91)
- ReCAD: Reinforcement Learning Enhanced Parametric CAD Model Generation with Vision-Language Models (AAAI 2026 · similarity 0.89)
- CAD-Coder: Text-to-CAD Generation with Chain-of-Thought and Geometric Reward (NeurIPS 2025 · similarity 0.89)
- CAD-Editor: A Locate-then-Infill Framework with Automated Training Data Synthesis for Text-Based CAD Editing (ICML 2025 · similarity 0.87)
- Generating CAD Code with Vision-Language Models for 3D Designs (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
