Metraipolis › archive › dc_ac1460f599bf

BlenderFORGE: Framework for Optimizing Reactive 3D-Graphics Editing Ability of MLLMs — arXiv:2611.10768

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.10768
License (as archived)
CC0-1.0
Archived
2026-11-10T18:41:12.300Z
SHA-256 of the archived text
d8d291c1dce2ddfdf078104a97ee3ddb461667894d67755929e42118489d339b
Stable URL of this page
https://metraipolis.com/documento/dc_ac1460f599bf

How to cite

BlenderFORGE: Framework for Optimizing Reactive 3D-Graphics Editing Ability of MLLMs — arXiv:2611.10768 — arXiv (en), https://arxiv.org/abs/2611.10768 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ac1460f599bf (Metraipolis archive dc_ac1460f599bf, sha256 d8d291c1dce2ddfdf078104a97ee3ddb461667894d67755929e42118489d339b)
BibTeX
@misc{dcac1460f599bf,
  title = {BlenderFORGE: Framework for Optimizing Reactive 3D-Graphics Editing Ability of MLLMs — arXiv:2611.10768},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.10768},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ac1460f599bf; sha256 d8d291c1dce2ddfdf078104a97ee3ddb461667894d67755929e42118489d339b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.10768. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Multimodal Large Language Models (MLLMs) have shown promise as interactive agents, yet precise, programmatic 3D graphics editing remains difficult to train because existing systems largely rely on closed-source model APIs and lack deterministic, state-verifiable execution environments. We present BlenderFORGE, a trainable data-environment-optimization framework for open-weight MLLM agents that edit explicit Blender scenes through Blender Python (bpy) code generation. BlenderFORGE targets four fundamental state-verifiable editing task families: object placement, shape-key editing, lighting adjustment, and material modification. It integrates three components: (1) a scalable perturbation-based data generation pipeline built from curated assets and large-scale 3D scene repositories; (2) an isolated Blender Sandbox for deterministic code execution, structured state export, and closed-loop reward computation; and (3) a two-stage training pipeline that uses offline teacher trajectories for cold-start Supervised Fine-Tuning (SFT) and sandbox-computed rewards for multi-turn tool-agentic Reinforcement Learning (AgentLoop-RL). Experiments across the four task families show that the resulting open-weight Qwen3-VL-8B and Qwen3.5-9B BlenderFORGE agents substantially improve over their base models and achieve competitive performance with several zero-shot proprietary MLLM baselines on selected metrics. Additional cross-benchmark evaluations and camera-reframing experiments provide evidence of transfer beyond the original training setting.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.10768 — 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 2026-11-10T18:37:37.504Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).