INDEX

Explanations

mental illusions and delusions

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

to

-2.11

 получила

-1.46

啻

-1.41

zzare

-1.32

 resear

-1.29

 حياة

-1.25

 adalah

-1.24

firstName

-1.23

就可以

-1.23

and

-1.22

POSITIVE LOGITS

ógy

1.36

大部分

1.27

 muuta

1.25

roveň

1.23

 boneka

1.21

satunya

1.20

Treating

1.17

 Pastikan

1.16

本店

1.16

żenie

1.16

Activations Density 0.032%