INDEX

Explanations

fascinating facts, trivia, bizarre, weird

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Direct

-0.08

 Demp

-0.08

sted

-0.08

 direct

-0.08

 dismant

-0.07

 interchangeable

-0.07

 Guidance

-0.07

artner

-0.07

 intim

-0.07

æĪĳçļĦ

-0.07

POSITIVE LOGITS

 trivia

0.21

 fascinating

0.20

 facts

0.20

 bizarre

0.19

 surprising

0.19

facts

0.18

 Facts

0.17

 weird

0.17

 Weird

0.17

 shocking

0.16

Activations Density 0.275%

No Known Activations