INDEX

Explanations

abilities are limited

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

 minimum

-0.09

indy

-0.09

 discour

-0.09

igm

-0.08

 simply

-0.08

wor

-0.08

dav

-0.08

avit

-0.08

 Minimum

-0.08

yme

-0.08

POSITIVE LOGITS

 limited

0.28

limited

0.24

 Limited

0.23

æľīéĻĲ

0.22

Limited

0.21

 limitations

0.20

 ÙħØŃØ¯ÙĪØ¯

0.20

 LIMITED

0.19

imited

0.18

 limitation

0.17

Activations Density 0.066%