INDEX

Explanations

D roads, dozer

The neuron activates on tokens containing numerals (Arabic digits), i.e. it spots numbers and numeric codes.

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

 comprador

-1.46

Oggi

-1.40

APPING

-1.38

 procurer

-1.38

اسية

-1.34

zył

-1.33

葎

-1.33

蔀

-1.30

 piese

-1.27

𞤢

-1.27

POSITIVE LOGITS

of

1.85

↵

1.68

 любых

1.44

];

1.30

 incluidas

1.30

 sidan

1.27

 vielen

1.26

es

1.25

Eacute

1.25

id

1.24

Activations Density 0.039%