INDEX

Explanations

malicious purposes, activities, code, actors

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

rib

-0.09

azen

-0.09

 IsNot

-0.09

 Nicola

-0.09

 Cycl

-0.08

 immersion

-0.08

stell

-0.08

akin

-0.08

 Twig

-0.08

.not

-0.08

POSITIVE LOGITS

ly

0.29

ness

0.23

 intent

0.18

LY

0.15

 actors

0.14

NESS

0.14

intent

0.14

ious

0.13

 Intent

0.13

 activity

0.12

Activations Density 0.025%