INDEX

Explanations

HTML and XML metadata

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

 мою

-2.09

というわけで

-1.96

 parede

-1.93

 ofreci

-1.92

);

-1.88

 کنیم

-1.85

を注文

-1.84

☫

-1.83



-1.83

 może

-1.83

POSITIVE LOGITS

and

2.98

in

2.67

to

2.16

Descripció

1.82

 usine

1.74

 along

1.73

鶲

1.72

minder

1.70

<>

1.69

cience

1.65

Activations Density 0.059%