INDEX

Explanations

place to store 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

Ì£

-0.10

 basket

-0.10

 baskets

-0.09

 prick

-0.09

 Basket

-0.09

 karak

-0.09

 pocket

-0.09

íĥķ

-0.08

 processing

-0.08

 dirt

-0.08

POSITIVE LOGITS

 storing

0.21

 storage

0.20

 stored

0.20

 stor

0.19

åŃĺ

0.17

 Storage

0.16

 store

0.16

storage

0.15

åĤ¨

0.15

(storage

0.15

Activations Density 0.011%