INDEX

Explanations

selected item or value

The neuron fires whenever the token “selected” (in any selection‐related context) appears.

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

</u>

-2.20

</em>

-1.57

following

-1.52

↵

-1.44

-1.41

-1.40

-1.38

 которые

-1.34

ero

-1.34

 принци

-1.31

POSITIVE LOGITS

any

1.68

ثم

1.59

 then

1.56

 anything

1.55

 вересня

1.50

 pembuatan

1.48

時間も

1.43

exemplar

1.42

 écologique

1.42

There

1.41

Activations Density 0.020%