mlsanityData to 14 Oct 2026

Recursive Code World Models: Building Complex Worlds through Recursive Scene Programs

arXiv 2609.11499 ↗First posted 10 Sep 2026cs.CVcited by 3

On the map: World models → Video-generation world models and simulators

Abstract (arXiv)

Code world models represent worlds as executable programs, but this representation alone does not determine how to construct a complex world. We introduce Recursive Code World Models (RCWM), a framework for reconstructing complex 3D worlds in code from a single reference image. RCWM couples a Recursive Scene Program (RSP) representation with a construction solver that recursively calls itself. An RSP represents the executable world as compositional scene code, while each solver call follows the same complete process: establish the whole, recursively reconstruct unresolved parts, and revisit the whole to refine their composition. This global-local-global recursion gives fine-scale structures their own perception-and-editing loops while preserving scene-wide geometry and relationships. Reference-aligned views propagate a shared camera projection across levels, while parent revisitation addresses boundaries, spatial relations, and shared errors that emerge after local refinement. A vision-language coding agent directly compares reference images with scene renders to guide refinement, recursive descent, and return. Across complex scenes, RCWM outperforms prior code-based image-to-scene reconstruction methods. Ablation studies further support the benefits of recursive construction and suggest that deeper calls can improve finer-scale reconstruction. RCWM provides a recursive construction principle for building complex executable worlds from visual evidence.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Video-generation world models and simulators, one of 11 directions in World models. Builds on 7 of its 11 references that are in the map. First posted on arXiv 10 Sep 2026.

Builds on

7 of 11 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026Code World Model: Coding Agent as World Brain26 Aug

Other references · 6

  1. 2024SceneCraft: An LLM Agent for Synthesizing 3D Scene as Blender Codecited by 134
  2. 2025Recursive Language Modelscited by 104
  3. 2026Vision-as-Inverse-Graphics Agent via Interleaved Multimodal Reasoningcited by 49
  4. 2026Thinking in Blender: Staged Executable Inverse Graphics with Vision-Language Modelscited by 10
  5. 2026WorldClaw: Agentic 3D Open-World Generation at Scalecited by 8
  6. 2026FuncRoom-Agent: Sequential Feed-Forward 3D Functional Indoor Scene Generationcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

Similar, not cited

Close in wording, no citation either way

  1. 2026Vision2Code: A Multi-Domain Benchmark for Evaluating Image-to-Code Generation
  2. 2026Code4Scene: Benchmarking Coding Agents for Constructing and Editing 3D Scenes
  3. 2026Code2Worlds: Empowering Coding LLMs for 4D World Generation
  4. 2026ExeWorldBench: Benchmarking Code-Based Executable 3D World Modeling
  5. 2026HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worlds
  6. 20263D Primitives are a Spatial Language for VLMs
  7. 2026VoxelCodeBench: Benchmarking 3D World Modeling Through Code Generation
  8. 2025IR3D-Bench: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering