INDEX

Explanations

UUIDs and identifiers

This neuron detects hexadecimal characters (digits and letters) within long identifier strings (e.g., UUIDs).

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

its

-1.41

ين

-1.33

to

-1.33

accompanied

-1.31

ن

-1.28

})

-1.27

usually

-1.25

%);

-1.24

ل

-1.24

是不

-1.23

POSITIVE LOGITS

ERVIEW

1.65

ätigung

1.63

餮

1.59

嚀

1.57

ulitis

1.51

 odstran

1.50

 možnosti

1.46

戾

1.44

 öde

1.44

uolo

1.44

Activations Density 0.018%