INDEX

Explanations

gain or hold power

The neuron responds to mentions of “power,” especially in contexts about assuming, holding, or rising to political power.

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

end

-1.63

 extrême

-1.60

🧌

-1.51

eſt

-1.48

 شناسی

-1.46

 sorpresas

-1.45

 nhưng

-1.42

 úplně

-1.41

מלי

-1.40

 véhic

-1.39

POSITIVE LOGITS

by

1.55

 município

1.48

 With

1.45

 with

1.45

 make

1.42

йки

1.35

 aimé

1.35

 bairro

1.30

 Until

1.30

тивная

1.30

Activations Density 0.016%