INDEX

Explanations

the

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

 Réponses

-0.80

 Roskov

-0.75

SharedCtor

-0.75

OGND

-0.75

 '\\;'

-0.68

findpost

-0.66

зулта

-0.65

 informée

-0.65

uxxxx

-0.63

вгений

-0.62

POSITIVE LOGITS

 elettrica

0.47

 render

0.47

 виправивши

0.46

 internet

0.43

 régionale

0.43

 nanoparticles

0.43

jani

0.42

tiver

0.41

 cables

0.41

@[

0.40

Activations Density 0.010%