INDEX

Explanations

much

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 much

-1.57

much

-1.56

Much

-1.50

 Much

-1.40

 MUCH

-1.18

 paljon

-0.99

 mucho

-0.96

 réfugi

-0.85

mucho

-0.83

mye

-0.81

POSITIVE LOGITS

 more

0.72

antaine

0.68

 information

0.68

 emphasis

0.67

 effort

0.66

 impact

0.65

 importance

0.65

 work

0.61

of

0.59

more

0.56

Activations Density 0.110%