INDEX

Explanations

section headers with numbers

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

𝙚

-1.36

but

-1.34

 unaweza

-1.32

"=>$

-1.26

 sentía

-1.23

 ekran

-1.20

 komfort

-1.20

 способом

-1.19

how

-1.18

 plakat

-1.16

POSITIVE LOGITS

 rigide

1.30

 aucune

1.25

選手の

1.23

foundland

1.21

 هیچ

1.20

什么的

1.18

 precin

1.15

":"+

1.14

ونه

1.14

ates

1.14

Activations Density 0.149%