INDEX

Explanations

phrases with specific or concrete nouns

The neuron fires on adjective tokens (POS tag “JJ”) in the text.

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

of

-1.65

and

-1.65

in

-1.52

 this

-1.51

on

-1.44

or

-1.41

at

-1.40

 also

-1.40

 that

-1.32

 where

-1.30

POSITIVE LOGITS

Zutaten

1.73

 abiye

1.55

 kemer

1.55

 kimse

1.52

 seksi

1.51

 refroid

1.51

 diki

1.45

쐈

1.45

 marchandises

1.45

 organik

1.45

Activations Density 0.008%