INDEX

Explanations

Punctuation and Code Snippets

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

parsedMessage

-1.01

 nakalista

-0.90

AccessorTable

-0.90

^(@)

-0.89

principalColumn

-0.87

aarrggbb

-0.86

 cherchés

-0.85

 Проци

-0.84

 ―――――

-0.83

Искәрмәләр

-0.82

POSITIVE LOGITS

‘

0.85

“

0.67

0.64

–

0.57

…

0.57

It

0.56

’

0.51

‘

0.50

<h4>

0.48

Activations Density 0.001%