INDEX

Explanations

change, silicon, mirror, sophisticated, using

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Много

1.65

斐

1.52

 король

1.49

แซ

1.47

mino

1.46

论

1.46

cement

1.44

 Чем

1.42

 Handicap

1.42

 života

1.42

POSITIVE LOGITS

 flip

1.48

det

1.42

adays

1.37

gangenheit

1.36

 trap

1.36

 unethical

1.36

uruan

1.35

 seien

1.34

 kontroll

1.34

DU

1.33

Activations Density 0.000%