INDEX

Explanations

expand, enhance, adapt

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

åĿĬ

-0.11

isco

-0.10

lluminate

-0.10

ÙĪØ§Øª

-0.09

#=

-0.09

 ÑĢÐ°ÑģÐº

-0.09

 escalated

-0.09

subs

-0.09

ãĥ³ãĥķ

-0.08

inflate

-0.08

POSITIVE LOGITS

 expansion

0.28

 expand

0.28

 divers

0.28

expand

0.26

 Expand

0.24

æī©

0.24

 expanding

0.22

 Expansion

0.22

Expand

0.22

Expansion

0.22

Activations Density 0.041%