INDEX

Explanations

the word message

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

.Cursors

-0.12

âĶ´

-0.10

 ä¸¶

-0.10

 Notify

-0.09

¶Į

-0.09

ResponseType

-0.09

 Calls

-0.09

_marshall

-0.09

.CustomButton

-0.09

ÅĻen

-0.09

POSITIVE LOGITS

 message

0.55

message

0.41

 Message

0.37

 messages

0.34

Message

0.34

.message

0.33

(message

0.32

_message

0.31

\tmessage

0.31

-message

0.29

Activations Density 0.123%