INDEX

Explanations

that followed by verb

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

 також

-1.99

 CONCEPTS

-1.90

ichtlich

-1.89

 myös

-1.80

<>();

-1.76

'+'

-1.76

(",")

-1.72

Controllo

-1.71

⸒

-1.70

甞

-1.68

POSITIVE LOGITS

2.55

or

2.30

 frankly

2.16

2.03

2.00

 includes

1.97

1.91

1.88

his

1.80

Activations Density 0.012%