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

Ã¸r

-0.27

åĵī

-0.26

orde

-0.25

è¢«æī§è¡Į

-0.25

_NB

-0.25

ors

-0.25

 prepar

-0.25

åĲİåį«

-0.24

éĤł

-0.24

 Ð´ÐµÐ½

-0.24

POSITIVE LOGITS

 wipe

0.28

wipe

0.26

ãģ¤ãģĭ

0.26

åıĺæĢģ

0.26

wal

0.26

atra

0.26

akeup

0.25

 wipes

0.24

 equivalent

0.24

anna

0.24

Activations Density 0.002%

No Known Activations

This feature has no known activations.