INDEX

Explanations

you know?

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wedge

1.74

 atrophy

1.59

 analysing

1.56

 investigating

1.55

 comparing

1.55

sna

1.50

 studying

1.50

 quantifying

1.47

hift

1.46

 discussing

1.46

POSITIVE LOGITS

TY

1.83

roveň

1.79

 klas

1.72

र्फ

1.68

เร

1.68

 cepat

1.67

렌

1.67

सेच

1.66

remen

1.65

仅

1.62

Activations Density 0.048%