INDEX

Explanations

normal, standard, regular, ordinary

The neuron fires on descriptive qualifier words—i.e. adjectives (like “normal,” “actual,” “standard,” “regular,” “expensive,” “neoprene”) used to qualify or describe nouns.

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

และ

-1.78

 différents

-1.60

 唯美

-1.50

<bos>

-1.46

 praticamente

-1.45



-1.41

་་

-1.41

炖

-1.38

があるので

-1.38

 vários

-1.38

POSITIVE LOGITS

 normal

1.80

 traditional

1.59

 Each

1.57

 regular

1.52

1.50

 ordinary

1.49

 обы

1.45

 standard

1.45

 heightened

1.43

normal

1.42

Activations Density 0.065%