INDEX

Explanations

ordering of steps or items

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

of

-1.51

not

-1.24

Prueba

-1.19

鼠

-1.14

 regarded

-1.13

앨

-1.09

OTO

-1.06

Rant

-1.06

nachron

-1.05

 that

-1.05

POSITIVE LOGITS

wego

1.28

记录

1.25

 участи

1.23

жают

1.17

MBOL

1.16

奬

1.14

Sebelumnya

1.14

大致

1.13

чните

1.13

 гле

1.12

Activations Density 0.052%