INDEX

Explanations

conjunctions and common phrases

understood primarily

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

consistent

0.50

ov

0.48

weekday

0.45

sharp

0.45

 Hungary

0.44

click

0.43

%，

0.43

ティブ

0.42

%。

0.41

POSITIVE LOGITS

辇

0.51

 محصولات

0.50

 glor

0.46

 ప్రభుత్వం

0.46

 massacre

0.45

 militar

0.44

缛

0.44

Productos

0.43

 Denis

0.42

 embell

0.42

Activations Density 0.001%