INDEX

Explanations

memory creation and sharing

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

bere

-0.11

aina

-0.10

bow

-0.09

AndGet

-0.09

Kaw

-0.08

 revel

-0.08

starter

-0.08

çª

-0.08

 Hobby

-0.08

 inherit

-0.08

POSITIVE LOGITS

 created

0.14

 creation

0.14

 shared

0.13

-making

0.13

 fond

0.12

-lasting

0.11

 Creation

0.11

et

0.11

 Shared

0.11

 Created

0.11

Activations Density 0.021%