INDEX

Explanations

with

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 człowieka

0.66

 oamenii

0.65

患者

0.64

 człowie

0.62

 사람들이

0.62

 člove

0.62

 मरीजों

0.61

お客様

0.61

 člov

0.60

人们

0.59

POSITIVE LOGITS

 disabilities

1.07

 whom

1.02

 Disabilities

0.97

 histories

0.96

whom

0.96

 impaired

0.92

 undergone

0.89

 impairments

0.80

 compromised

0.80

 जिनकी

0.75

Activations Density 0.074%