INDEX

Explanations

how to perform an action

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

Negative Logits

ocab

-0.10

unm

-0.09

 Daly

-0.09

 considered

-0.09

ÑĢÐ°Ð±

-0.09

lass

-0.09

 Rosenstein

-0.08

lick

-0.08

 Clyde

-0.08

Fah

-0.08

POSITIVE LOGITS

 proceed

0.23

 approached

0.19

 Proceed

0.18

 proceeding

0.17

 proceeded

0.17

 approach

0.17

 handle

0.16

 handled

0.16

 proceeds

0.15

 handles

0.14

Activations Density 0.098%

No Known Activations