INDEX

Explanations

Roth names and Roth followed by specific words

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

Regardless

-0.75

 themes

-0.75

Though

-0.73

ｳ

-0.72

ाम

-0.71

schrank

-0.70

魔兽

-0.70

 menambah

-0.68

പ്പ

-0.67

mentable

-0.67

POSITIVE LOGITS

child

0.91

suf

0.89

Verte

0.87

 IRON

0.84

Roth

0.84

magne

0.80

milan

0.77

 وزارة

0.75

Radial

0.74

hamma

0.74

Activations Density 0.030%