INDEX

Explanations

separate and independent

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

双方

-0.84

리아

-0.82

ükemmel

-0.80

Beskrivning

-0.80

 ensuring

-0.76

 ironically

-0.75

タウン

-0.75

During

-0.73

 during

-0.71

Autoritní

-0.71

POSITIVE LOGITS

 separate

2.59

separate

2.17

 separately

2.16

Separate

2.14

 seperate

2.00

 Separate

1.97

 SEPAR

1.82

 separat

1.78

 отдельно

1.72

 sépar

1.70

Activations Density 0.118%