INDEX

Explanations

Thor

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

egend

-0.31

å¥½åĲĹ

-0.27

enco

-0.26

olare

-0.25

 olacaÄŁÄ±

-0.25

éħĨ

-0.24

 agency

-0.24

MakeRange

-0.24

çļĦåģļæ³ķ

-0.24

eness

-0.24

POSITIVE LOGITS

ripp

0.31

é¡¿

0.29

RI

0.28

rie

0.26

Sie

0.26

 ÐºÐ¾Ð¼

0.25

My

0.25

ynec

0.25

ym

0.25

steen

0.25

Activations Density 0.103%