INDEX

Explanations

positive adjectives

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

الدراسه

-0.49

وفة

-0.47

 Wiktionnaire

-0.47

SCRIPTION

-0.46

âtel

-0.45

ucket

-0.44

gnant

-0.44

 BorderSide

-0.44

лению

-0.44

OUNTS

-0.44

POSITIVE LOGITS

 deal

0.77

 number

0.77

 many

0.66

 amount

0.63

 Anzahl

0.61

number

0.60

esModule

0.59

many

0.58

 DEAL

0.57

 feat

0.57

Activations Density 0.086%