INDEX

Explanations

purposes and reasons

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

for

1.35

 voor

1.28

for

1.24

圧力

1.22

對於

1.21

 высокая

1.20

jala

1.19

huana

1.18

voor

1.16

ettle

1.15

POSITIVE LOGITS

 sake

2.58

 purposes

2.39

 reasons

2.05

 Sake

1.93

 Purposes

1.80

 posterity

1.77

 starters

1.73

 betterment

1.59

 eternity

1.59

 raisons

1.58

Activations Density 1.081%