INDEX

Explanations

independent of

The neuron fires on occurrences of the word “independently,” especially when used in the phrase “independently of/from.”

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

あります

-1.62

 Ähn

-1.60

 todavía

-1.42

Livro

-1.39

 downright

-1.39

};

-1.39

 Onde

-1.33

}{*}{}

-1.33

Still

-1.32

同じく

-1.32

POSITIVE LOGITS

 that

2.36

it

1.65

but

1.63

abstractmethod

1.62

prende

1.55

，

1.55

you

1.54

ΙΣ

1.48

 katk

1.46

lierung

1.42

Activations Density 0.031%