INDEX

Explanations

car descriptions

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

 Cadillac

-0.12

 Jeep

-0.12

 automobiles

-0.11

 Ferrari

-0.10

 Rentals

-0.10

 jeep

-0.10

 rentals

-0.10

 Chevrolet

-0.09

 Porsche

-0.09

 Chevy

-0.09

POSITIVE LOGITS

 Kelley

0.21

 Lemon

0.16

 Kelly

0.16

 Edmund

0.15

Hag

0.15

Car

0.14

 Carm

0.14

CAR

0.14

 lemon

0.14

VIN

0.13

Activations Density 0.063%