INDEX

Explanations

specific context in multiple languages

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

"She

-0.29

“;She

-0.29

 ÐºÐ¾ÑĤÐ¾ÑĢÑĥÑİ

-0.22

 ÐºÐ¾ÑĤÐ¾ÑĢÐ°Ñı

-0.21

ÑİÑīÐ°Ñı

-0.21

 quella

-0.20

 ktÃ³ra

-0.20

 Ø§ÙĦØ±ÙħØ²ÙĬØ©

-0.19

 à¤Ĺà¤Ī

-0.19

ovanÃ¡

-0.19

POSITIVE LOGITS

 celui

0.34

 ÐºÐ¾ÑĤÐ¾ÑĢÑĭÐ¹

0.29

 quello

0.28

 kterÃ½

0.26

izzato

0.26

 ÐºÐ¾ÑĤÐ¾ÑĢÐ¾Ð¼

0.26

Ð°Ð½Ð½ÑĭÐ¹

0.26

Ð»ÐµÐ½Ð½ÑĭÐ¹

0.25

owany

0.24

ativo

0.24

Activations Density 0.476%