INDEX

Explanations

electronic controls

prevention, paused, direction, elements

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

 मैती

0.57

gheatmap

0.53

 הס

0.52

 дать

0.51

TimeDict

0.51

 независимо

0.50

 ομά

0.50

 अपॉन

0.49

insula

0.48

 طلبه

0.47

POSITIVE LOGITS

":

0.54

കു

0.49

hi

0.46

zny

0.45

 Gran

0.43

";

0.42

zab

0.42

xe

0.41

 leth

0.40

is

0.39

Activations Density 0.004%