INDEX

Explanations

text related to research and academic papers, especially in the medical and social science domains

this neuron activates for 'imageNamed'

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

Negative Logits

et

-0.49

Халык

-0.48

zkod

-0.48

těte

-0.48

asca

-0.48

 للاسماء

-0.47

 perfectos

-0.46

“

-0.45

les

-0.45

“

-0.44

POSITIVE LOGITS

 imageNamed

0.68

MENAFN

0.62

)";

0.59

!")

0.59

twimg

0.58

getSystemService

0.57

pidou

0.56

)");

0.55

".

0.54

$",

0.54

Activations Density 3.678%