INDEX

Explanations

inj

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

 needed

-0.28

éĽ¾

-0.28

 Capt

-0.26

ãģ£ãģ¦ãģĦãģı

-0.25

åĹ·

-0.25

eph

-0.25

éĴı

-0.24

 warranted

-0.24

fog

-0.24

izing

-0.23

POSITIVE LOGITS

edList

0.34

æĹģ

0.32

åĤį

0.29

-earth

0.28

ä¸Ģä»½

0.26

 immune

0.26

å¹´ä¹ĭ

0.25

ä»½

0.25

 earth

0.25

_cmos

0.24

Activations Density 0.003%