INDEX

Explanations

code keywords and punctuation

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

TestMethod

-0.76

 sév

-0.72

ITTLE

-0.69

 rodzin

-0.68

ração

-0.68

Padding

-0.67

timi

-0.66

 receivable

-0.66

HAM

-0.66

 Wrath

-0.66

POSITIVE LOGITS

osen

0.77

Loft

0.71

bigr

0.70

estacks

0.69

дку

0.68

sī

0.67

Халы

0.67

 Thon

0.67

 Polsek

0.66

edo

0.66

Activations Density 0.020%