INDEX

Explanations

numbers and items

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

poj

-0.10

 MetroFramework

-0.10

Pixels

-0.09

ë°°

-0.09

 nÃ¸

-0.09

ippets

-0.09

/respond

-0.09

EMU

-0.09

 portion

-0.08

 WaitForSeconds

-0.08

POSITIVE LOGITS

 main

0.15

 total

0.15

 entr

0.12

 Entr

0.11

entr

0.11

 TOTAL

0.11

total

0.10

 possible

0.10

 primary

0.10

Activations Density 0.132%