INDEX

Explanations

false appearance or disguise

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 unseen

-0.11

igin

-0.10

 tong

-0.10

Vu

-0.10

âĢ²

-0.10

uai

-0.10

 invisible

-0.09

 Murdoch

-0.09

adb

-0.09

 Pavilion

-0.09

POSITIVE LOGITS

 faÃ§

0.14

facade

0.14

 guise

0.13

 vene

0.12

 cover

0.12

 mask

0.12

 disguise

0.11

/fa

0.11

 outward

0.11

 fronts

0.11

Activations Density 0.072%