INDEX

Explanations

vulnerability

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

옥

1.88

कुंठ

1.84

 قادر

1.83

우

1.81

수의

1.78

팀

1.77

توان

1.77

밭

1.76

 اعلی

1.75

)\|_{

1.74

POSITIVE LOGITS

 scorching

2.25

 cours

2.22

 Maintenant

2.17

 contraception

2.15

 Ahora

2.14

 cession

2.13

 peanut

2.11

 slight

2.08

 entrance

2.05

 maintenant

2.03

Activations Density 0.081%