INDEX

Explanations

anatomy or errors

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 rear

-0.87

 Rear

-0.84

rear

-0.78

 REAR

-0.70

Rear

-0.68

 larmes

-0.62

 ouvriers

-0.59

GEBURTS

-0.57

الحياه

-0.57

 trasera

-0.57

POSITIVE LOGITS

IsMutable

0.63

UpInside

0.56

mybatisplus

0.55

zat

0.54

ammer

0.53

 defaultstate

0.53

gating

0.50

NOPQRST

0.49

facing

0.48

ranging

0.48

Activations Density 0.031%