INDEX

Explanations

the other killers. killed others

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

OffsetTable

-0.10

.rar

-0.09

hua

-0.09

çĿĢ

-0.09

@Web

-0.09

abus

-0.09

-div

-0.08

erland

-0.08

Orm

-0.08

GuidId

-0.08

POSITIVE LOGITS

 original

0.17

 kill

0.16

 killing

0.16

 kills

0.14

 originally

0.14

kill

0.13

 person

0.13

æĿĢ

0.12

 killed

0.12

 alive

0.12

Activations Density 0.044%