INDEX

Explanations

expressed or intensified states

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

люби

-0.91

rahman

-0.87

relenting

-0.83

teau

-0.82

sortBy

-0.80

чення

-0.79

'\

-0.78

mais

-0.77

ree

-0.76

ˆ

-0.76

POSITIVE LOGITS

 expressed

1.47

 stronger

1.37

 misplaced

1.34

源

1.22

expressed

1.20

 outweigh

1.13

wan

1.11

 manifested

1.10

旺

1.09

 override

1.06

Activations Density 0.056%