Gemma 4 E4b gemma-4-e4b

google · gemma

unranked — no quality signals yet

params
8B
arch
dense
license
apache-2.0 open weights
released
Mar 2026
downloads/30d
550.1k

# architecture

attn shared GQA 8q/2kv feed-forward (dense) every parameter, every token out ×42 layers 8B — no routing
layers
42
d_model
2560
heads
8q / 2kv
head dim
256
vocab
262144
family
gemma4_text

signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.

# why ranked

Overall: provisional — score 22.8 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_intelligence 0.70 22.8
bench_composite 0.30 —

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

OGM score · last 44 days
OGM score over 44 days
overall rank (up is better)
Overall rank over 44 days

# benchmarks

benchmarkscoresourcedate
AA Intelligence Index 8.9 Artificial Analysis —

# where to run

no known API providers — download the weights from Hugging Face and run them yourself.

source aliases
aa
gemma-4-e4b