INDEX

Explanations

closing quotes followed by words

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 oluyor

-1.73

讶

-1.66

 geliyor

-1.58

 başlad

-1.55

 هذه

-1.55

 sağlar

-1.55

 ihtiyac

-1.50

および

-1.47

 bulunabilir

-1.46

</h2>

-1.45

POSITIVE LOGITS

老者

1.78

⎠

1.44

んでいた

1.42

if

1.38

rógeno

1.37



1.36

 when

1.34

 asisten

1.33

.-"

1.33

 that

1.31

Activations Density 0.048%