INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æĸ¹å¼ıè¿Ľè¡Į

-0.29

atk

-0.28

çļĦæłĩåĩĨ

-0.27

amp

-0.26

åĸĥ

-0.25

 similar

-0.25

 subsystem

-0.25

 form

-0.25

ç±»ä¼¼çļĦ

-0.24

avar

-0.24

POSITIVE LOGITS

moid

0.27

`"]↵

0.26

irie

0.26

"]=="

0.25

elder

0.25

"],"

0.25

ei

0.24

å®¹

0.24

è½¦åºĵ

0.24

")+

0.24

Activations Density 0.001%

No Known Activations

This feature has no known activations.