INDEX

Explanations

listing items

The neuron flags tokens that are numeric or contain digits (e.g. “234,” “185,” “109,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 OGSÅ

-0.96

槩

-0.87

 zatem

-0.86

 conceito

-0.85

 vín

-0.85

 quita

-0.84

 znan

-0.84

 aluno

-0.83

 firmy

-0.83

museo

-0.82

POSITIVE LOGITS

 lieutenant

0.84

 diversions

0.83

 Confeder

0.80

 wanderer

0.79

ведений

0.79

 persen

0.78

ĩnh

0.78

ceo

0.77

 zprá

0.76

 pixabay

0.75

Activations Density 0.003%