INDEX

Explanations

directional adjectives followed by places

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

åĲĦç§įåĲĦæł·

-0.22

 Reich

-0.21

imon

-0.21

runner

-0.19

clip

-0.19

 clip

-0.19

 dome

-0.19

ench

-0.19

 Walls

-0.18

 Streets

-0.18

POSITIVE LOGITS

æĿĳåºĦ

0.22

ÙĩØ±

0.20

åĩºåıĳ

0.20

most

0.20

åĮºåĿĹ

0.19

éĢļè½¦

0.19

å±±åĮº

0.19

å¼Ģå±Ģ

0.19

æ¾Ħæ¸ħ

0.19

æ·±å¤Ħ

0.19

Activations Density 0.047%