INDEX

Explanations

recovery

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 recover

-0.98

 recovers

-0.93

 recovering

-0.89

 recovered

-0.84

recover

-0.72

 recovery

-0.71

recovered

-0.65

 Recover

-0.64

 myſelf

-0.62

 RECOVERY

-0.62

POSITIVE LOGITS

TagMode

0.60

#+#

0.58

Enfin

0.55

KommentareTeilen

0.54

 ***!

0.54

Spoljašnje

0.54

uable

0.52

Rüyada

0.50

enumi

0.50

 bois

0.49

Activations Density 0.045%