INDEX

Explanations

the ..."}, noun."}, phrase."}, noun."}, noun."

The neuron is mainly detecting occurrences of the definite article “the” (and, more broadly, very 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

such

-1.42

should

-1.41

튿

-1.39

心想

-1.37

 وهذا

-1.33

 ledit

-1.32

 mismo

-1.31

 outras

-1.28

moved

-1.27

だけは

-1.26

POSITIVE LOGITS

of

1.80

 årene

1.49

 interested

1.46

 gustado

1.42

 fjor

1.41

 Vordergrund

1.39

ärä

1.37

 ของ

1.35

why

1.32

 nogen

1.30

Activations Density 0.009%