INDEX

Explanations

pronouns followed by verbs

This neuron responds strongly to high-frequency function words—especially personal pronouns (he, she, they, you) and common articles like “the.”

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

now

-2.14

 also

-1.84

 then

-1.74

or

-1.68

and

-1.55

</strong>

-1.45

elerik

-1.45

derdag

-1.43

 такие

-1.40

 cristales

-1.39

POSITIVE LOGITS

1.66

НИМА

1.54

誒

1.51

 HAVING

1.50

1.38

картинка

1.34

 NÃO

1.34

 bahwa

1.31

 trône

1.30

Activations Density 0.045%