INDEX

Explanations

first

implied government backing

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

 fordul

0.43

pur

0.42

үр

0.41

 Straf

0.41

遄

0.39

Finale

0.39

 Aula

0.39

айл

0.38

 euismod

0.38

 المتع

0.38

POSITIVE LOGITS

টিও

0.46

 startling

0.42

 painstaking

0.40

 REPORTER

0.39

 নেতৃত্ব

0.39

بہ

0.39

 softened

0.38

ocked

0.37

ંગી

0.37

 স্বপ্নের

0.37

Activations Density 0.009%