INDEX

Explanations

worth the cost or risk

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

醞

-1.40

越來

-1.34

 défendre

-1.27

安裝

-1.27

 mehrfach

-1.24

知っている

-1.22

惴

-1.18

 verstor

-1.18

 многое

-1.16

彼は

-1.16

POSITIVE LOGITS

imagenes

1.30

 itinéraire

1.28

[$

1.27

:^{

1.27

 آنچه

1.26

 zonder

1.25

 fresa

1.25

家的

1.20

 necessárias

1.20

ностью

1.19

Activations Density 0.016%