INDEX

Explanations

introduces contrast or drawback

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

 feels

1.74

 rests

1.55

ba

1.52

 nobody

1.47

 felt

1.47

 each

1.46

 they

1.44

 nothing

1.42

 infus

1.40

 devotes

1.40

POSITIVE LOGITS

ว้

2.05

روع

2.03

termilk

1.97

lerini

1.95

เภท

1.94

هلاك

1.91

لول

1.86

โยชน์

1.86

ternut

1.85

μφωνα

1.84

Activations Density 0.080%