Metraipolis › archive › dc_efaace44d15d
3DCodeEdit: Benchmarking Instruction-Guided 3D Program Editing — arXiv:2610.38231
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.38231
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-31T06:16:45.036Z
- SHA-256 of the archived text
7a62d6bacf8b786be60df432d85c442885f4937b415bdaee156aed869aa14cec- Stable URL of this page
https://metraipolis.com/documento/dc_efaace44d15d
How to cite
3DCodeEdit: Benchmarking Instruction-Guided 3D Program Editing — arXiv:2610.38231 — arXiv (en), https://arxiv.org/abs/2610.38231 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_efaace44d15d (Metraipolis archive dc_efaace44d15d, sha256 7a62d6bacf8b786be60df432d85c442885f4937b415bdaee156aed869aa14cec)
BibTeX
@misc{dcefaace44d15d,
title = {3DCodeEdit: Benchmarking Instruction-Guided 3D Program Editing — arXiv:2610.38231},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.38231},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_efaace44d15d; sha256 7a62d6bacf8b786be60df432d85c442885f4937b415bdaee156aed869aa14cec},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.38231. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Language models increasingly create 3D content through code, with practical workflows often relying on editing existing programs rather than regenerating them, for speed and precise control. Such edits require understanding what a program builds and foreseeing how a change reshapes it, yet prior benchmarks mostly evaluate generation or simple edits against a reference. In this paper, we introduce 3DCodeEdit, a benchmark of 2,592 instruction-guided editing tasks over 108 Blender programs that construct objects entirely in code. Its edits range from parameter to structural changes, style redesign, and scene composition, and are judged from renders against human-reviewed checklists. Across 17 models, the average success rate is only 37%, and even the strongest models complete fewer than 60% of structural and style edits. As models become more capable, their failures shift to geometric foresight, where the code runs but builds the wrong shape. We therefore study whether multi-turn feedback can compensate for this missing foresight. Feedback from a separate reviewer model that describes how the renders differ from the request helps most, more than maximum reasoning effort or an agent harness, and can let a 27B model surpass the single-shot rate of a 2.8T-parameter model. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL) ; Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR). Pagina del lavoro: https://arxiv.org/abs/2610.38231 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-31T06:07:23.327Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).