INDEX

Explanations

percentage changes

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

 bricolaje

-1.72

Cuando

-1.55

ждает

-1.49

 همه

-1.45

และ

-1.41

 joyería

-1.39

 fieltro

-1.39

ierter

-1.38

 cuarzo

-1.37

 personalización

-1.35

POSITIVE LOGITS

in

2.42

 while

1.97

to

1.72

 from

1.59

at

1.52

 during

1.51

on

1.45

 within

1.39

 both

1.38

for

1.33

Activations Density 0.023%