INDEX

Explanations

imp

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

sup

-0.44

Sup

-0.39

å£ĵ

-0.39

sup

-0.38

åİĭ

-0.38

-sup

-0.36

 compression

-0.35

_sup

-0.34

 suppression

-0.34

 Compression

-0.32

POSITIVE LOGITS

å¢ŀè¿Ľ

0.34

åĩıå°ĳäºĨ

0.31

æıĲé«ĺäºĨ

0.30

 improves

0.28

lify

0.28

oney

0.27

éĻįä½İäºĨ

0.27

ä¿ĥè¿Ľ

0.27

ä¿ĥè¿ĽäºĨ

0.27

 enhances

0.26

Activations Density 0.782%