Coding Agent Is Good As World Simulator

arXiv 2026 55.9 method, system

TLDR

An agentic framework that builds physics-based world models by generating and iteratively refining executable simulation code, outperforming video-based models.

Reasoning

The paper presents a novel approach using code generation for physically plausible world models, with clear methodology and experimental results. However, the abstract lacks details on the scale of experiments and does not specify real-world benchmarks, limiting assessment of generalizability.

Read-first score

Read-first score 55.9, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 42.

Recency 6%

100

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

Citation impact 18%

75.4

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.75353428

Methodology quality 18%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=analysis,experiment,result

Topical relevance 29%

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 18%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=code

Citation velocity 12%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FoundationFrontierBridgeMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 274.

Keyword Scores

world model

world simulator

world dynamics prediction

interactive world model

generative world model

model-based reinforcement learning world model

video world model

Deep Analysis

Innovations

Agentic framework that constructs physics-based world models through executable simulation code, combining planning, code generation, visual review, and physics analysis agents.
Iterative revision of simulation code based on visual and physical feedback to enforce physical constraints and match prompt requirements.
Demonstrates superiority over video-based world models in physical accuracy, instruction fidelity, and visual quality.

Methodology

The framework coordinates four agents: a planning agent converts natural language prompts into structured scene plans, a code agent implements the plan as executable simulation code, a visual review agent provides visual feedback, and a physics analysis agent checks physical consistency. The code is iteratively revised based on feedback until the simulation matches prompt requirements and physical constraints.

Key Results

The proposed framework outperforms advanced video-based models in physical accuracy, instruction fidelity, and visual quality, and can be applied to driving simulation and embodied robot tasks.

Resources

Paper