INDEX

Explanations

accessing object attributes/methods

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

igay

1.70

oelect

1.45

deut

1.43

itare

1.41

ifed

1.41

 שהוא

1.39

ңыз

1.38

izont

1.38

 włos

1.37

μμα

1.37

POSITIVE LOGITS

use

1.67

 sends

1.63

her

1.59

 uses

1.56

 использует

1.48

saw

1.46

 send

1.42

.$$

1.40

 pays

1.38

はや

1.37

Activations Density 0.001%