INDEX

Explanations

new stages and directions

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

 econó

-1.32

 actuellement

-1.25

 gutes

-1.17

 alguna

-1.16

 róż

-1.16

 izquier

-1.07

 actuales

-1.04

余計

-1.04

ズキ

-1.04

再來

-1.03

POSITIVE LOGITS

 with

1.44

for

1.39

 than

1.38

in

1.31

 possibilities

1.30

新的

1.29

 затем

1.23

>−</

1.23

new

1.18

、

1.16

Activations Density 0.053%