INDEX

Explanations

part of speech

The neuron identifies content‐bearing words—salient nouns, verbs, and other meaningful tokens—while ignoring common function words.

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

 första

-1.83

 fyra

-1.67

 skolan

-1.56

 någon

-1.55

 использования

-1.53

については

-1.52

 två

-1.51

 empez

-1.49

 åren

-1.47

 flera

-1.45

POSITIVE LOGITS

1.64

1.52

may

1.36

of

1.31

 either

1.18

 each

1.14

 information

1.12

 under

1.12

 without

1.10

</b>

1.09

Activations Density 0.052%