INDEX

Explanations

numbered stages or categories

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

ของ

-1.35

 نیز

-1.23

 again

-1.23

この

-1.21

 expres

-1.17

 также

-1.14

entown

-1.09

furter

-1.09

return

-1.07

Procedimento

-1.07

POSITIVE LOGITS

 first

2.09

 pertama

1.63

แรก

1.51

 пър

1.41

 první

1.39

 ersten

1.38

 första

1.34

 eerste

1.32

 первая

1.30

 pierws

1.27

Activations Density 0.036%