# Num\_experts\_per\_tok for MoE models

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fnum-experts-per-tok-for-moe-models%2F85405%3E%5C
**Category:** Models\
**Created:** [May 6, 2024, 11:08am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fnum-experts-per-tok-for-moe-models%2F85405 "2024-05-06T11:08:03Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![LuckiestOne](https://avatars.discourse-cdn.com/v4/letter/l/2bfe46/32.png) [@LuckiestOne](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FLuckiestOne)%5C
**Post date:** [May 6, 2024, 11:08am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fnum-experts-per-tok-for-moe-models%2F85405%2F1 "2024-05-06T11:08:03Z")

</div>

The description for this param (for example in the mixtral config) states

```
        The number of experts to root per-token, can be also interpreted as the `top-p` routing

```

I am not quite sure I understand this correctly. As this is an integer (set to two in the mixtral case) - why is it called top-p instead of top-k routing?
