INDEX

Explanations

always

almost always

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

जनिक

0.46

രു

0.37

 semelh

0.37

حديث

0.36

 potencialmente

0.35

िम

0.34

Ḍ

0.34

роятно

0.33

ارد

0.32

տր

0.32

POSITIVE LOGITS

 always

4.94

always

4.72

 ALWAYS

4.72

 всегда

4.59

 sempre

4.56

 Always

4.53

Always

4.50

 siempre

4.47

 завжди

4.34

 zawsze

4.28

Activations Density 0.505%