INDEX

Explanations

activities or subjects

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

atcher

-0.10

 enrich

-0.10

vid

-0.09

istrovstvÃŃ

-0.09

 ÑıÐºÑĥ

-0.08

.generated

-0.08

 pert

-0.08

croll

-0.08

aze

-0.08

POSITIVE LOGITS

 capable

0.13

 Äĳá»ĥ

0.12

 ÑĩÑĤÐ¾Ð±Ñĭ

0.12

ä¸ºäºĨ

0.11

 inorder

0.10

 surrounding

0.10

ynom

0.10

 that

0.10

elpers

0.09

ìĿ´ë¹Ħ

0.09

Activations Density 0.164%