INDEX

Explanations

are followed by an adjective or past participle

are followed by descriptions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ع

0.86

0.85

ו

0.82

 yake

0.73

 έχει

0.72

 auffi

0.72

an

0.71

でも

0.71

л

0.70

ם

0.70

POSITIVE LOGITS

ari

0.78

0.69

oles

0.67

ate

0.66

wolves

0.65

ats

0.64

 themselves

0.64

eli

0.64

ants

0.63

are

0.62

Activations Density 0.362%