Gemma 3
Google DeepMind — Gemma
partial detail 20 method fields published
Evaluation
Multilingual benchmark evaluation
performance of pre-trained models on multilingual tasks with in-context learning (multi-shot prompting) on MGSM, Global-MMLU-Lite, WMT24++, FLoRes, XQuAD, ECLeKTic, IndicGenBench and XOR QA
How it was run
Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.