Weevaluate ourmodel onbothsynthetic andin-the-wild Internet imagery and demonstrate its advantages in terms of generalization and accuracy.

The need to analyze graphs is ubiquitous across various fields, from social networks tobiological research and recommendation systems.

Comprehensiveevaluationson MSQA andMSNN highlight thelimitations ofexisting vision-language models and underscore the importance of handling multi-modal interleaved inputs and situation modeling.

Our benchmark comprises three existing datasets with expanded annotations to supportmulti-taskevaluation.