INDEX

Explanations

can followed by verbs

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

()!=

-1.18

)}

-1.12

TagHelpers

-1.08

 быть

-1.07

>\<

-1.06

અ

-1.06

 recomendaciones

-1.05

ásticas

-1.03

 funktionieren

-1.03

백

-1.02

POSITIVE LOGITS

 also

2.75

 either

2.17

 even

1.87

use

1.82

 expect

1.68

 find

1.64

either

1.62

 easily

1.57

 then

1.44

see

1.43

Activations Density 0.057%