Guess the Unified Model: How Much Can We Recover from Generated Images?
This work addresses the underexplored problem of attributing unified model-generated images to their source model, providing a foundation for transparency and auditing in generative image pipelines.
The paper investigates whether images generated by different unified models can be attributed to their source model, finding near-perfect attribution accuracy with ~20K images per model, minimal impact from corruptions, and that semantic content aids but does not dominate separability.
With unified model-generated images now widespread online, attributing their model of origin offers a path toward transparency and deeper insight into the characteristic behaviors of individual models. Prior work has explored provenance in LLM-generated text, diffusion model images, and datasets, but the separability of unified model-generated images remains an underexplored area. We address this gap by examining separability across corruption, domains, and prompt languages using images generated by seven unified models. We show that model attribution is highly feasible as our model achieves near-perfect accuracy with around 20K images per model. Corruptions and structural perturbations have only a modest effect on attribution performance, and cross-domain generalization reveals that semantic content contributes to separability but is not the dominant signal. Finally, we observe that for most models, prompt language attribution is around chance levels, suggesting minimal language-specific visual signatures. These findings highlight consistent model-specific visual characteristics in unified models outputs and open new directions for tracing and auditing generative image pipelines.