Visual Credit Audit for Multimodal Spatial Reasoning
Quick summary
arXiv:2607.27069v2 Announce Type: cross Abstract: Closed yes/no spatial benchmarks can reward a correct answer even when the image adds little support beyond no-image contexts. Under a fixed forced-choice interface, Visual Credit Audit (VCA) separates two estimands: whether the benchmark image gives the model's declared decision more support than text-only and blank controls, and whether the model responds to relation-specific visual evidence. The first audit is training- and label-free and does not require an answer flip. Applying labels yields dependence-credited correctness (D-CC); on corre
Key takeaways
- arXiv:2607.27069v2 Announce Type: cross Abstract: Closed yes/no spatial benchmarks can reward a correct answer even when the image adds little support beyond no-image contexts.
- Under a fixed forced-choice interface, Visual Credit Audit (VCA) separates two estimands: whether the benchmark image gives the model's declared decision more support than text-only and blank controls, and whether the model responds to relation-specific visual evidence.
- The first audit is training- and label-free and does not require an answer flip.
Why it matters
“Visual Credit Audit for Multimodal Spatial Reasoning” should be evaluated beyond branding and benchmark scores. Its practical importance will emerge in task accuracy, latency, unit cost, safety and integration with real workflows.