INDEX

Explanations

arising from use

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

νονται

-0.92

Alzheimer

-0.91

 tiennent

-0.90

cés

-0.89

 oslo

-0.88

osť

-0.87

 cultura

-0.87

そこは

-0.86

러나

-0.85

 vagas

-0.85

POSITIVE LOGITS

or

1.56

 this

1.33

any

1.13

use

1.05

 activities

0.99

 arise

0.99

 transactions

0.98

 arises

0.98

การ

0.98

 ANYTHING

0.98

Activations Density 0.016%