INDEX

Explanations

function followed by ()

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

 côn

-1.59



-1.56

 bénéficiaire

-1.52



-1.50

雱

-1.44

 první

-1.43

ೢ

-1.42

 aguj

-1.41

 vijf

-1.37

 paño

-1.37

POSITIVE LOGITS

 with

2.05

in

1.80

1.79

 from

1.71

</strong>

1.61

,”

1.57

on

1.55

at

1.48

1.36

”

1.36

Activations Density 0.006%