INDEX

Explanations

but rather

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

'gc

-0.15

iyon

-0.10

abouts

-0.10

à¹Īà¸ģ

-0.10

ãĥ¬ãĥ³

-0.09

.override

-0.09

idge

-0.09

eda

-0.09

WARE

-0.08

 Ð¡ÐµÑĢ

-0.08

POSITIVE LOGITS

 rather

0.31

 sondern

0.24

 Rather

0.24

rather

0.23

nor

0.22

 instead

0.21

Rather

0.20

 Ø¨ÙĦÚ©Ùĩ

0.19

 spÃŃÅ¡e

0.17

 Instead

0.17

Activations Density 0.027%