INDEX

Explanations

generics and coding constructs

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

 have

-1.98

 their

-1.90

 your

-1.87

 several

-1.86

 these

-1.84

 because

-1.78

 there

-1.73

 some

-1.70

などを

-1.65

 first

-1.63

POSITIVE LOGITS

1.53

1.49

ecznie

1.49

 Перейти

1.45

künfte

1.44

大多

1.41

tować

1.38

ubert

1.35

1.34

稍微

1.32

Activations Density 0.001%