INDEX

Explanations

connect with clients

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

Negative Logits

Vul

-0.09

UPPORTED

-0.09

Eld

-0.09

etail

-0.09

fox

-0.08

spinner

-0.08

wei

-0.08

 doPost

-0.08

çļ

-0.08

co

-0.08

POSITIVE LOGITS

 clients

0.26

 Clients

0.24

 client

0.24

clients

0.20

-client

0.20

Client

0.19

Clients

0.19

 Client

0.19

 CLIENT

0.18

client

0.18

Activations Density 0.068%

No Known Activations