INDEX

Explanations

and or leading to result

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

1.53

and

1.46

และ

1.39

и

1.37

and

1.27

:,

1.27

 และ

1.24

but

1.22

và

1.21

 এবং

1.20

POSITIVE LOGITS

 whatnot

2.51

 other

1.99

 hatta

1.91

 sebagainya

1.90

 bahkan

1.83

 consequently

1.77

 consequent

1.77

 tentunya

1.77

 somit

1.73

 autres

1.68

Activations Density 2.700%