INDEX

Explanations

end of sentence or clause

The neuron fires on major content words—especially nouns and pronouns—rather than on 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

 horaires

-1.25

 créer

-1.25

ـ

-1.24

 dando

-1.23

 zwar

-1.22

嘴唇

-1.21

씰

-1.20

 mouvements

-1.19

 ajouter

-1.18

 zoude

-1.16

POSITIVE LOGITS

 just

2.08

 only

1.84

 give

1.61

 when

1.59

at

1.59

 makes

1.57

 gives

1.55

or

1.55

 there

1.54

now

1.52

Activations Density 0.033%