Interactive diagram of the Gemma 4 E2B model: inputs flow through encoders into a stack of decoder layers that predict the next token, with panels on attention windows and effective versus total parameters.
The whole model, input to output
Click any block above for a plain-English note.
Inside one decoder layer
Click a stage above to see what it does.
Attention window
Effective vs total parameters
2.3B effective, does the thinking5B total stored
2.3B
The gap is the PLE tables. They are big to store but only used for cheap lookups, so the compute cost matches a 2.3B model. Total footprint runs under 1.5 GB, small enough for a phone.