INDEX

Explanations

abstract states and issues

negative consequences or outcomes

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

cknowled

0.43

ǎi

0.42

 এগ

0.41

 Rural

0.41

 এইসব

0.41

 তাঁরা

0.39

正则表达式

0.39

 använder

0.38

they

0.38

 käyttää

0.38

POSITIVE LOGITS

 havoc

0.64

 bagi

0.63

 meille

0.62

带来

0.61

 لنا

0.61

 untold

0.60

遐

0.60

帶來

0.59

 στους

0.58

 आम्हाला

0.57

Activations Density 0.155%