INDEX

Explanations

communicating ideas and information

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

ích

-0.95

 掛け

-0.93

ците

-0.92

 inférieur

-0.91

евич

-0.90

hender

-0.89

就好

-0.88

Begriffsklä

-0.86

buch

-0.86

crizione

-0.85

POSITIVE LOGITS

 ideas

2.09

 information

1.92

 what

1.79

 thoughts

1.61

ideas

1.46

 Ideas

1.42

to

1.41

how

1.41

 message

1.40

 messages

1.37

Activations Density 0.039%