INDEX

Explanations

assignments and numbered items

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

animous

-0.99

思議

-0.87

hipster

-0.85

 Worry

-0.84

 initiating

-0.81

建設

-0.81

復興

-0.78

аб

-0.75

kadi

-0.74

dasarkan

-0.74

POSITIVE LOGITS

 två

1.09

 fyra

1.02

nº

0.98

 sixth

0.96

 cuatro

0.96

４

0.95

eight

0.94

0.93

 week

0.93

１

0.92

Activations Density 0.188%