INDEX

Explanations

directions

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

zed

-0.08

åĩºä¾Ĩ

-0.08

ÙĦÙĬØª

-0.07

 mukaan

-0.07

åĩºåĬ¨

-0.07

Ð½Ð¾ÑģÑĤÑĮ

-0.07

itys

-0.07

åįĹåİ¿

-0.07

edReader

-0.07

 dazu

-0.07

POSITIVE LOGITS

æĸ¹åĲĳ

0.12

æĸ¹åĲĳçļĦ

0.11

 direction

0.10

wards

0.09

 Richtung

0.09

ward

0.09

æĸ¹åĲĳåıĳå±ķ

0.09

æĸ¹åĲĳåĴĮ

0.08

-direction

0.08

direction

0.08

Activations Density 0.020%

No Known Activations