INDEX

Explanations

IP addresses and numerical data points

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

No Comments

Negative Logits

########.

-0.70

 оригіналу

-0.66

исленность

-0.57

司

-0.50

sh

-0.50

roll

-0.49

WithIOException

-0.49

司

-0.49

 crossorigin

-0.48

كويكب

-0.47

POSITIVE LOGITS

ſelf

0.58

DockStyle

0.57

 laiton

0.56

Efq

0.55

 Houſe

0.55

 uſed

0.54

 Theſe

0.54

 Diſ

0.54

 Chrif

0.53

 uſe

0.53

Activations Density 1.077%