INDEX

Explanations

directional locations

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-1.58

 erwäh

-1.53

 regalías

-1.44

 inmiddels

-1.40

 jiné

-1.38

 entwickelte

-1.35

 erzä

-1.34

 ehemal

-1.34

 entwick

-1.30

diğini

-1.29

POSITIVE LOGITS

of

2.56

 ideen

1.41

 essentielle

1.34

 visuelle

1.32

 spirituelle

1.31

 flott

1.27

 кало

1.26

 Desember

1.24

 ihrer

1.23

тори

1.22

Activations Density 0.068%