INDEX

Explanations

and followed by action phrases

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

ématiques

-1.72

>--}}

-1.71

↵↵

-1.63

-1.62

Insulin

-1.61

and

-1.59

久しぶりの

-1.56

スポンサーリンク

-1.47

.«

-1.45

 gelangt

-1.42

POSITIVE LOGITS

is

1.91

 lusso

1.61

 either

1.60

 further

1.56

 between

1.55

this

1.54

 Handlungen

1.50

 pronta

1.49

 According

1.48

閨

1.48

Activations Density 0.099%