INDEX

Explanations

positive outcomes and contrast

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

veras

-0.86

 Ebers

-0.85

subpackage

-0.83

 adicionais

-0.81

hringer

-0.81

 ajustado

-0.81

ˍ

-0.80

onOptions

-0.78

renti

-0.77

thee

-0.77

POSITIVE LOGITS

 also

0.87

 glories

0.83

 exalt

0.81

Forced

0.80

 Geset

0.80

 promoção

0.78

 forced

0.78

 triumphs

0.78

 reward

0.77

 joys

0.76

Activations Density 0.173%