INDEX

Explanations

the word "neural"

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

No Comments

Negative Logits

WPS

-0.69

WO

-0.68

Rolf

-0.68

 Sabrina

-0.67

keo

-0.67

 Tatiana

-0.67

 Cosmo

-0.67

nski

-0.66

/*

-0.64

 Winthrop

-0.63

POSITIVE LOGITS

Neural

1.35

neural

1.32

 Neural

1.26

 neural

1.23

 muun

0.87

شاركة

0.76

 Judea

0.75

jména

0.73

 Emery

0.73

gridx

0.72

Activations Density 0.002%