INDEX

Explanations

number of employees

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

 Individual

-0.10

 backlog

-0.09

 weighing

-0.09

/npm

-0.09

ä¸ļ

-0.09

enha

-0.09

enums

-0.09

åĨł

-0.09

weigh

-0.08

olland

-0.08

POSITIVE LOGITS

 numbers

0.27

 size

0.25

 roster

0.22

numbers

0.21

 number

0.20

size

0.20

 sizes

0.20

 composition

0.19

-size

0.19

 ÑĩÐ¸Ñģ

0.18

Activations Density 0.161%