INDEX

Explanations

assessing thoroughness

purification

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

 Elat

0.51

 عادی

0.50

谐

0.45

DIRECTORY

0.44

婴

0.44

Auch

0.43

СП

0.43

Hej

0.43

 অপার

0.42

Thirty

0.42

POSITIVE LOGITS

ro

0.46

 prepping

0.44

ítja

0.44

 purification

0.43

 trin

0.43

 prepares

0.43

 tarp

0.43

 darken

0.43

ಗ್ರ

0.43

та

0.42

Activations Density 0.000%