What it measures

Evaluates generative video and 3D world models on physical consistency, camera control, temporal coherence, and instruction adherence.

Reading the measurement

Study what information the model receives and which inference the answer requires. A question may mix recall, visual perception, deduction and search. Use task breakdowns to find the mechanism behind a headline gain.

Published evaluation

The source’s static and dynamic world-generation composites remain separate. These are generative-video and world-model results on the original 0–100 scale.

Where this fits in an evaluation suite

Ask what the task actually requires: recall, deduction, search, or a mixture. Separate closed-book reasoning from retrieval and tool-assisted problem solving.

Start with a capability hypothesis, then choose a second test with a different failure mode. Keep the checkpoint, reasoning effort, agent tools, sampling policy, and scoring procedure fixed when measuring a training intervention. Record failures and missing outputs in the denominator.

A reproducible comparison

The source’s static and dynamic world-generation composites remain separate. These are generative-video and world-model results on the original 0–100 scale.

Publish the exact run configuration, item coverage, score units, and available uncertainty. If a source supplies only an aggregate or a single run, preserve that limit. A new model name or a rounded score alone does not establish a capability improvement.

PRIMARY EVIDENCE & FURTHER READING

  1. WorldScore: A Unified Evaluation Benchmark for World Generation — published result table ↗

    Haoyi Duan, Hong-Xing Yu, Sirui Chen, Li Fei-Fei, Jiajun Wu · 2025

Reviewed 2026-10-26. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.