INDEX

Explanations

hexadecimal numbers and byte values

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

正常的

-0.99

Vind

-0.92

普通的

-0.84

 Kläger

-0.81

鞦

-0.80

ニム

-0.79

 revés

-0.77

AUF

-0.76

σαι

-0.75

ostar

-0.75

POSITIVE LOGITS

pytest

0.86

status

0.83

hver

0.81

command

0.81

 drier

0.79

fec

0.78

 others

0.77

 interactiva

0.77

 Marion

0.76

 lille

0.76

Activations Density 0.005%