INDEX

Explanations

X is/was Y

This neuron responds to nouns that introduce the central element or choice under discussion—words like “question,” “option,” “factor,” “exception,” or the “best bet.”

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

 başlad

-1.45

来週

-1.40

明日は

-1.39

 obuv

-1.34

 alapján

-1.31

 totul

-1.28

凄く

-1.27

になりますが

-1.27

朝から

-1.27

só

-1.25

POSITIVE LOGITS

is

4.41

was

2.94

are

2.91

คือ

1.83

 είναι

1.78

は

1.66

的是

1.59

1.50

つは

1.49

—

1.36

Activations Density 0.190%