Benchmark

VQAv2 (val)

VQAv2 is a balanced Visual Question Answering dataset containing open-ended questions about images that require understanding of vision, language, and…

Modality
multimodal
Categories
multimodal, reasoning, image_to_text, language, vision
Openness
unknown
Source
llm_stats
Reported scores
3

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
Gemma 3 12BGoogle0.7162025-03-12evidence
Gemma 3 27BGoogle0.712025-03-12evidence
Gemma 3 4BGoogle0.6242025-03-12evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.