INDEX

Explanations

settlement

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

pears

-0.62

 prisonniers

-0.59

 propOrder

-0.57

udière

-0.57

 esclavos

-0.56

PETS

-0.56

 délic

-0.55

meau

-0.55

uxley

-0.54

DOCTYPE

-0.54

POSITIVE LOGITS

 patterns

0.96

 pattern

0.87

 PATTERN

0.72

patterns

0.65

 Patterns

0.64

AndEndTag

0.57

Patterns

0.57

 history

0.56

 patrones

0.56

 Pattern

0.54

Activations Density 0.004%