INDEX

Explanations

closing punctuation and brackets

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

to

-1.64

in

-1.54

頭痛

-1.42

–

-1.29

and

-1.25

-1.20

ircut

-1.19

 Despite

-1.18

 أخرى

-1.16

To

-1.14

POSITIVE LOGITS

CLUYE

1.67

 capucha

1.64

(`

1.59

alties

1.57

tamaan

1.52

 fordi

1.52

 manguera

1.52

饪

1.52

itize

1.49

 tratamientos

1.49

Activations Density 0.005%