INDEX

Explanations

health

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 BoxDecoration

-0.83

SourceChecksum

-0.73

quelize

-0.70

 مرئيه

-0.67

scryfall

-0.66

Personendaten

-0.65

DECREF

-0.65

rungsseite

-0.64

 disambiguazione

-0.61

 חיצוניים

-0.60

POSITIVE LOGITS

response

0.47

hold

0.43

who

0.42

security

0.42

carved

0.42

command

0.41

ọt

0.41

aby

0.40

order

0.40

before

0.40

Activations Density 0.022%