Research map / Large language models / Agents, code and math
Towards High-Fidelity CAD Generation via LLM-Driven Program Generation and Text-Based B-Rep Primitive Grounding
Jiahao Li, Qingwang Zhang, Qiuyu Chen, Guozhan Qiu, Yunzhong Lou, Xiangdong Zhou
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 |
|---|---|---|---|
| add RL stage after SFT (ablation) | FutureCAD · FutureCAD Advanced | Invalid rate: 42.67 -> 8.29 | Improved |
| add BRepGround text-based primitive selection (ablation) | FutureCAD · FutureCAD Advanced | Invalid rate: 8.29 -> 1.01 | Improved |
| add BRepGround text-based primitive selection (ablation) | FutureCAD · FutureCAD Advanced | Mean CD: 76.01 -> 71.39 | Improved |
| add BRepGround text-based primitive selection (ablation) | FutureCAD · FutureCAD Advanced | Median CD: 38.50 -> 31.12 | Improved |
| add BRepGround text-based primitive selection (ablation) | FutureCAD · FutureCAD Advanced | R-F1: 82.22 -> 85.35 | Hurt |
| add RL and BRepGround to SFT only (ablation) | FutureCAD · FutureCAD Advanced | R-F1: 91.44 -> 85.35 | Hurt |
| add RL stage after SFT (ablation) | FutureCAD · FutureCAD Advanced | G-F1: 79.82 -> 80.36 | Improved |
| use BRepGround cross-attention grounding instead of Late Fusion | BRepGround · B-Rep grounding | F1: 43.32 -> 50.23 | Improved |
| use BRepGround instead of CLIP-DE | BRepGround · B-Rep grounding | mAP: 57.53 -> 63.07 | Improved |
| FutureCAD vs CAD-LLaMA extended baseline | FutureCAD · FutureCAD Advanced | Invalid rate: 46.60 -> 1.01 | Improved |
| FutureCAD vs CADFusion on standard subset | FutureCAD · FutureCAD Standard | G-F1: 91.91 -> 90.39 | Hurt |
Closest papers by meaning
- CAD-Llama: Leveraging Large Language Models for Computer-Aided Design Parametric 3D Model Generation (CVPR 2025 · similarity 0.91)
- CAD-Coder: Text-to-CAD Generation with Chain-of-Thought and Geometric Reward (NeurIPS 2025 · similarity 0.91)
- Text2CAD: Generating Sequential CAD Designs from Beginner-to-Expert Level Text Prompts (NeurIPS 2024 · similarity 0.90)
- FlexCAD: Unified and Versatile Controllable CAD Generation with Fine-tuned Large Language Models (ICLR 2025 · similarity 0.89)
- CAD-Editor: A Locate-then-Infill Framework with Automated Training Data Synthesis for Text-Based CAD Editing (ICML 2025 · similarity 0.88)
- From Intent to Execution: Multimodal Chain-of-Thought Reinforcement Learning for Precise CAD Code Generation (AAAI 2026 · similarity 0.88)
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.
