INDEX

Explanations

indicating someone

tell someone

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

学生的

0.44

本身的

0.39

 Anything

0.38

自身的

0.37

 suas

0.36

 multe

0.36

 liczb

0.36

 vielfält

0.36

 இந்தியாவின்

0.36

它的

0.36

POSITIVE LOGITS

 whoever

1.34

 someone

1.28

 somebody

1.14

 anyone

1.10

 alguien

1.08

someone

1.07

Whoever

1.05

 qualcuno

1.04

 Whoever

1.02

 alguém

0.98

Activations Density 0.606%