INDEX

Explanations

original noun/source

original intent or meaning

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 beher

0.57

 previo

0.56

ยอม

0.56

 düz

0.55

 prac

0.51

 hibernation

0.50

 sanitizer

0.49

ಲಾಗ

0.49

 dibujar

0.49

 oscilloscope

0.49

POSITIVE LOGITS

original

1.56

 original

1.40

 originale

1.39

 Original

1.34

 ORIGINAL

1.28

Original

1.27

 ориги

1.20

 orijinal

1.19

 оригі

1.17

 оригина

1.17

Activations Density 0.222%