INDEX

Explanations

command

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

 Fairview

0.86

 lens

0.80

 Fair

0.78

 Toronto

0.78

 fair

0.78

Sunny

0.77

 Mauer

0.77

liers

0.75

 Tris

0.75

 matur

0.75

POSITIVE LOGITS

 Commands

3.23

 Command

3.03

Command

3.01

command

2.95

 COMMAND

2.88

commands

2.85

COMMAND

2.77

Commands

2.74

 command

2.72

 commands

2.67

Activations Density 0.009%