Just came across this coding benchmark: SciCode Artificialanalysis.ai reports a ranking which contradicts the feeling we've towards those models in real life coding. Is Gemma 4 really that good, or a benchmarking issue? EDIT: The contribution of this benchmark to the Intelligence index of artificialanalysis.ai: Full Intelligence Index v4.1 weights: GDPval-AA v2: 20% Terminal-Bench 2.1: 16% τ³-Bench Banking: 14% Humanity's Last Exam: 12% AA-Omniscience Accuracy: 8% SciCode: 8% GPQA: 6% ...