CANVAS 2D UNAVAILABLE
This demo draws a log-zoom scale tunnel of parameter counts on a canvas.
Idea in one line: 7B means roughly 7,000,000,000 learned numbers (weights).
Weight file size = parameter count × bits per weight ÷ 8. Precision changes
the file, never the count. MoE stores all experts but computes with a subset.
Allocating 7,000,000,000 scalars

PROTRAILBLAZER

What 7B Means · Parameter Scale Tunnel
TOTAL PARAMS7,000,000,000
ACTIVE / TOKEN= TOTAL
BITS / PARAM16
WEIGHT FILE14.00 GB
RELATIVE SCALE×7.00 VS 1B
ARCHDENSE
COUNT SWEEP0
B = BILLION LEARNED WEIGHTS
NOT FACTS · TOKENS · NEURONS · LINES OF CODE
Count is size, not quality. Capability depends on architecture, data, training, quantization, and the serving stack. Weight file = params × bits ÷ 8 (exact, 1 GB = 10⁹ bytes); running a model needs more (KV cache, activations). The MoE split (8 experts, top-2) and the training ×3 to ×8 band are ILLUSTRATIVE.