INDEX

Explanations

statement of principle

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

1.19

盔

0.87

 reins

0.83

 uphe

0.81

 affected

0.80

 repaired

0.80

(=

0.79

 repair

0.79

(?)

0.78

if

0.76

POSITIVE LOGITS

 nomor

1.32

 menghilangkan

1.20

 persyaratan

1.19

 protokol

1.18

 keperluan

1.18

 layanan

1.17

 mandato

1.16

 parametros

1.15

 jellem

1.14

 lantai

1.14

Activations Density 0.019%