INDEX

Explanations

location words following the

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

in

-1.82

titure

-1.57

 leído

-1.32

 conocía

-1.32

 suceder

-1.31

 sebuah

-1.29

 trató

-1.27

 cambiado

-1.25

 more

-1.24

POSITIVE LOGITS

</h2>

1.55

 каждой

1.48

%;

1.44

;}

1.44

 каждого

1.43

the

1.38

 lesquelles

1.38

 برای

1.38

 которых

1.35

 если

1.34

Activations Density 0.066%