OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling

Bowei Liu, Xinchen Zhang, Xuhuan Li, Tianlin Zhang, Kaian Jiang et al.|October 15, 2026arXiv

Key Takeaway

Graded, component-level verification with pairwise comparison significantly outperforms binary verification for evaluating complex multimodal outputs, especially in structured reasoning tasks where compositional correctness matters.

Summary

This paper introduces OmniVerifier-C1, a visual verifier that moves beyond simple True/False judgments to provide detailed, component-level scores for evaluating generated images.

evaluation reasoning multimodal

Key Terms

graded-verification sub-question-decomposition pairwise-verification world-modeling decoupled-rl-training