INDEX

Explanations

invisible or stealth

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

мых

-0.96

mbal

-0.84

 знаний

-0.81

Vien

-0.81

estrogen

-0.81

Interactive

-0.81

磊

-0.77

 کند

-0.74

umburg

-0.72

윈

-0.71

POSITIVE LOGITS

 invisible

3.75

 Invisible

3.17

invisible

3.17

Invisible

2.84

 invis

2.59

visibility

2.52

INVISIBLE

2.22

 visibility

2.11

 Invis

2.03

 cloak

1.91

Activations Density 0.026%