Active parameters
Also: activated parameters, active params
In a mixture-of-experts model, the small part of the model actually used for each word. It decides speed; the whole model still has to fit in memory.
The fuller definition
In a mixture-of-experts model, the active parameters are the subset actually used to produce each token. They set the model's speed, while the total parameter count sets how much memory it needs.
- Total
All experts in memory
- Router
Picks a few per token
- Active
What is read per token
A simplified relationship, not a screenshot or a complete design.
A concrete example
Qwen3.6-35B-A3B has 35 billion parameters in total and about 3 billion active per token.
The distinction that matters
Use total parameters to check fit and active parameters to estimate speed.
Related terms
Mixture of experts (MoE)
A model built from several specialist sub-networks, with only a few switched on for each piece of text. It saves work without losing size.
FoundationsParameter
One of the billions of adjustable numbers inside a model. Training sets them; using the model does not change them.
Hardware & hostingMemory bandwidth
How fast data moves out of memory, in gigabytes per second. Think of it as the width of a loading door: wider means the model writes words faster.
Where it is taught
How models learn from examplesSource and further reading
Hugging Face · LLM courseThe definition and example are written for this library. The link gives background, official documentation or legal authority as labelled. Legal scope depends on the jurisdiction and current text; product features depend on release and settings.