INDEX

Explanations

run by

This neuron detects mentions of an organization or service being “run/operated by” (i.e. who manages or governs something).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

in

-1.77

</em>

-1.63

 before

-1.59

 אחרים

-1.43

vább

-1.40

喲

-1.32

-1.29

一秒

-1.26

 anderer

-1.25

 timmar

-1.24

POSITIVE LOGITS

mathrm

1.80

 obligé

1.61

</h1>

1.55

 mismas

1.48

也

1.48

 ‘‘

1.47

あー

1.47

1.46

 mêmes

1.45

」，

1.42

Activations Density 0.055%