INDEX

Explanations

is not aware

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ï½Ĵ

-0.10

 stranger

-0.09

azi

-0.08

 foul

-0.08

 jealous

-0.08

ult

-0.08

 capable

-0.08

æĮ¯

-0.08

æĿ¥èĩª

-0.08

beat

-0.08

POSITIVE LOGITS

 unsus

0.22

 vulnerability

0.21

 vulnerabilities

0.19

 vulner

0.19

 vulnerable

0.19

 unaware

0.17

 Vulner

0.16

 defense

0.15

 defence

0.14

hap

0.14

Activations Density 0.117%