INDEX

Explanations

many of

The neuron activates on quantifier words expressing large amounts or extent (e.g. “many,” “so,” “much”), i.e. terms that denote magnitude.

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

 even

-1.56

-1.41

 extremely

-1.40

even

-1.38

 beispielsweise

-1.37

すごく

-1.31

ed

-1.30

THERE

-1.30

 significantly

-1.29

POSITIVE LOGITS

of

2.56

tahu

1.48

lijden

1.45

 máme

1.37

옳

1.33

 наш

1.32

ของ

1.31

ыр

1.29

曖昧さ回避

1.29

 oportun

1.28

Activations Density 0.014%