INDEX

Explanations

individual parts of a whole

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

 puente

-0.86

BindingResult

-0.84

 Características

-0.84

 initData

-0.83

リーナ

-0.82

 meist

-0.82

 oltre

-0.80

 całym

-0.79

 trasero

-0.79

 énergé

-0.77

POSITIVE LOGITS

 each

2.83

 individual

2.81

individual

2.34

Individual

2.28

 einzelnen

2.27

Each

2.03

 Individual

1.93

 einzelne

1.93

 individually

1.81

 poszczegól

1.80

Activations Density 0.077%