INDEX

Explanations

automatically stored

describing harmful acts

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

uq

0.50

 שלנו

0.47

 לא

0.46

éral

0.45

0.43

uk

0.43

aque

0.42

 রান্না

0.41

ל

0.41

ăț

0.41

POSITIVE LOGITS

 Kentucky

0.47

 Dana

0.46

 Florida

0.44

 Dakota

0.44

Hoc

0.44

 Citation

0.44

 supplemental

0.43

 Montana

0.43

 Hauptstadt

0.43

 Indiana

0.42

Activations Density 0.000%