INDEX

Explanations

applicable law

The neuron activates on content‐heavy nouns—especially proper names, place names, and technical or domain‐specific terms.

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

 icke

-1.54

olge

-1.52

ں

-1.48

bedien

-1.43

 olika

-1.42

seende

-1.41

 gangen

-1.40

 wiele

-1.39

dekor

-1.38

quelize

-1.36

POSITIVE LOGITS

by

1.87

 would

1.82

 there

1.49

There

1.49

for

1.45

all

1.41

an

1.41

Would

1.39

in

1.36

of

1.36

Activations Density 0.139%