INDEX

Explanations

human

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

 beings

-0.47

 being

-0.38

being

-0.36

eten

-0.35

-Ãªtre

-0.33

-being

-0.32

ream

-0.29

 souls

-0.29

Being

-0.28

åĥ®

-0.28

POSITIVE LOGITS

resources

0.42

èµĦæºĲ

0.41

_resources

0.38

-resources

0.38

resource

0.38

 Resource

0.36

 resource

0.36

è³ĩæºĲ

0.35

 resources

0.35

 Resources

0.35

Activations Density 0.017%