WorldClaw: Agentic 3D Open-World Generation at Scale

arXiv 2026 29.4 method, system

TLDR

WorldClaw is an agentic coarse-to-fine framework for generating large-scale 3D open worlds from text with coherent terrain and editable assets.

Reasoning

The paper presents a novel agentic pipeline for coherent 3D scene generation, but lacks quantitative evaluation or real-world benchmarks, and its focus is on open-world generation rather than world models or dynamics prediction.

Read-first score

Read-first score 29.4, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 0.

Recency 8%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Topical relevance 42%

20.6

Matches configured research keywords against title, abstract, tags, and analysis text. matched=4

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=none

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 137.

Keyword Scores

world model

world simulator

generative world model

interactive world model

video world model

world dynamics prediction

model-based reinforcement learning world model

Resources

Paper