Model

Gemma 3 4B

Google

deepmind.google

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
BLiMPPI93.4accuracyN/AIndependentmediumarxiv.org
GPQA Diamond23.2accuracyN/AIndependenthighepoch.ai
OTel66.2scoreN/AIndependentmediumarxiv.org
tinyqabenchmark_core-en86.5exact_matchN/AIndependentmediumarxiv.org

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
PatchBench18.2hncsN/AIndependentmediumarxiv.org

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
UkrMeme0.3hf1N/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…