INDEX

Explanations

potential negative outcomes

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 gestión

2.51

 generar

2.46

 dédié

2.37

ेक्ट

2.29

 compét

2.27

 generan

2.24

 Ettha

2.24

 haces

2.22

 actualización

2.20

 korte

2.20

POSITIVE LOGITS

1.89

1.77

In

1.72

It

1.65

In

1.64

'\

1.59

And

1.58

Blast

1.57

Oklahoma

1.57

 This

1.55

Activations Density 0.116%