INDEX

Explanations

look

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

 houſe

-0.78

 Czechoslovakia

-0.75

 Moslem

-0.74

 Cubans

-0.71

 itſelf

-0.70

 fevere

-0.69

 Houſe

-0.69

 communism

-0.68

 ſever

-0.68

 miſ

-0.67

POSITIVE LOGITS

 resourceCulture

0.63

GEBURTSDATUM

0.60

EDEFAULT

0.59

WebServlet

0.58

 under

0.55

 समीक्षक

0.54

WriteAttribute

0.54

SharedDtor

0.53

parsedMessage

0.52

ⓧ

0.52

Activations Density 0.066%