INDEX

Explanations

words related to scientific research and reporting

punctuation marks

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

Negative Logits

Datuak

-1.09

 kasarigan

-1.01

 ―――――

-0.98

 resourceCulture

-0.98

enderror

-0.96

 Jefus

-0.95

 myſelf

-0.92

rungsseite

-0.91

 Мексичка

-0.91

 becauſe

-0.89

POSITIVE LOGITS

0.65

0.56

0.54

0.52

“

0.50

0.49

0.48

↵↵

0.47

Activations Density 7.871%