List your company

Visual Question Answering on ViRL39K

50AccuracyThinking model version
-211.52538.5Apr 23, 2026

Evaluation results

MethodLinks
Thinking model versionCategory=Broader STEM Topics, Count=22026.04505142,904.5
Thinking model versionCategory=(GradeSchool) Geometric, Count=262026.0434.62233.53,482.15
Thinking model versionCategory=Overall, Count=722026.0420.83379.13,317.29
Thinking model versionCategory=Spatial Reasoning, Count=62026.0416.67434.53,398.67
Thinking model versionCategory=(GradeSchool) Non-Geo Math, Count=212026.0414.29302.193,640.62
Thinking model versionCategory=Tables/Diagrams/Charts, Count=122026.048.33760.832,097.92
Thinking model versionCategory=(GradeSchool) Science, Count=32026.040408.334,096
Thinking model versionCategory=Social Science, Count=22026.0404444,096