INDEX

Explanations

card's magnetic strip

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

 Hollow

-0.10

 ransom

-0.09

 Lana

-0.09

 Creed

-0.09

 Candid

-0.09

ocache

-0.08

sensor

-0.08

 Passport

-0.08

ad

-0.08

 License

-0.08

POSITIVE LOGITS

SIM

0.25

sim

0.23

 card

0.21

SIM

0.20

_SIM

0.19

Sim

0.18

 chip

0.18

 sims

0.17

 cards

0.16

 reader

0.16

Activations Density 0.037%