INDEX

Explanations

{ data structure }

This neuron detects tokens containing digits (numerical values or identifiers with numbers).

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

 when

-2.02

or

-1.91

 used

-1.76

 during

-1.70

if

-1.70

 before

-1.63

 your

-1.57

out

-1.56

can

-1.55

at

-1.54

POSITIVE LOGITS

 célè

1.73

よな

1.68

耷

1.67

皷

1.66

 siè

1.66

 diriger

1.64

盉

1.60

Faça

1.59

 bleus

1.55

 fenô

1.54

Activations Density 0.038%