INDEX

Explanations

select best most important

most important and relevant elements

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

 женщина

0.44

 another

0.43

 வேற

0.41

ə

0.40

 অনুরূপ

0.40

 മറ്റൊരു

0.39

 utilisée

0.39

 อ่า

0.38

 አንድ

0.37

 aucun

0.36

POSITIVE LOGITS

那些

0.83

 aquellos

0.77

 those

0.77

 ceux

0.76

those

0.75

 quelli

0.73

 ones

0.72

 diejenigen

0.72

 select

0.70

 outliers

0.69

Activations Density 0.301%