INDEX

Explanations

4

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

BIO

-0.27

olvers

-0.27

bows

-0.26

å®ļåĲĳ

-0.26

redd

-0.25

ockets

-0.25

\Action

-0.24

 bikes

-0.24

è¹²

-0.24

uck

-0.23

POSITIVE LOGITS

Talking

0.26

WH

0.26

isphere

0.25

ä½Ĩæĺ¯æĪĳ

0.25

Vir

0.25

epar

0.24

disk

0.24

åįıä¼ļä¼ļåĳĺ

0.24

 inconsistency

0.23

åľĪ

0.23

Activations Density 0.045%