INDEX

Explanations

Lightweight and or even underwear

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

 apparel

-0.14

 Clothes

-0.14

 footwear

-0.12

 Apparel

-0.12

shirt

-0.12

 clothes

-0.12

 clothing

-0.11

 attire

-0.11

 Shoe

-0.11

 Shirt

-0.11

POSITIVE LOGITS

 jumps

0.18

cor

0.15

bod

0.15

 chem

0.14

 boilers

0.14

 pencil

0.13

 suit

0.13

tun

0.13

rom

0.12

 boiler

0.12

Activations Density 0.055%