INDEX

Explanations

strike through

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

⇀

-0.80

 PEAR

-0.75

IPT

-0.74

 khăn

-0.74

warna

-0.74

 Profesional

-0.73

·

-0.71

 LOCK

-0.70

 الموس

-0.68

Klo

-0.68

POSITIVE LOGITS

 strike

2.02

 underline

1.88

strike

1.67

Strike

1.65

 line

1.59

 crossed

1.51

underline

1.50

 Strike

1.49

 underlined

1.44

 underlines

1.44

Activations Density 0.028%