Benchmarks

8 results across 8 benchmarks

Benchmark results for Gemma 4 E2B Instruct
BenchmarkScoreSource
BIG-Bench Extra Hard Original — micro averageReported as: thinking mode21.9%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on BIG-Bench Extra Hard Original — micro average (opens in a new tab)Micro-average accuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on BIG-Bench Extra Hard Original — micro average
GPQA DiamondReported as: thinking mode43.4%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on Graduate-Level Google-Proof Q&A Diamond (opens in a new tab)GPQA Diamond AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on Graduate-Level Google-Proof Q&A Diamond
IFBench OriginalReported as: thinking mode38.0%arxiv.org for Gemma 4 E2B Instruct (thinking mode) on IFBench Original (opens in a new tab)AccuracyReported Technical reportGoogle DeepMindSelf-reportedRecord for Gemma 4 E2B Instruct on IFBench Original
IFEval OriginalReported as: thinking mode94.6%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on IFEval Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on IFEval Original
MATH-Vision OriginalReported as: thinking mode52.4%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on MATH-Vision Original
MedXpertQA MMReported as: thinking mode23.5%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on MedXpertQA MM (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on MedXpertQA MM
MMLU-Pro OriginalReported as: thinking mode60.0%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on MMLU-Pro Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on MMLU-Pro Original
SciCode OriginalReported as: thinking mode21.0%arxiv.org for Gemma 4 E2B Instruct (thinking mode) on SciCode Original (opens in a new tab)AccuracyReported Technical reportGoogle DeepMindSelf-reportedRecord for Gemma 4 E2B Instruct on SciCode Original