INDEX

Explanations

wondering about how, what, where, if

wondering and questions

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

how

0.75

how

0.72

 কীভাবে

0.69

 cómo

0.64

 어떻게

0.63

 nasıl

0.61

 bagaimana

0.60

cómo

0.60

 कैसे

0.59

如何

0.59

POSITIVE LOGITS

 wonder

1.06

wonder

0.92

 Wonder

0.91

Wonder

0.86

 wondered

0.85

 wonders

0.78

 aloud

0.75

 wondering

0.75

 Wunder

0.71

 Wonders

0.68

Activations Density 0.017%