INDEX

Explanations

closing 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

if

-1.10

 diverse

-0.87

例えば

-0.85

 actually

-0.83

 just

-0.82

 enhances

-0.81

 includes

-0.81

 combines

-0.80

 COMPOUNDS

-0.79

 incorporate

-0.78

POSITIVE LOGITS

it

1.10

toDouble

1.06

anii

0.91

怎

0.90

 такая

0.89

 serene

0.88

inematics

0.85

ſp

0.84

 экспери

0.84

胰

0.84

Activations Density 0.006%