INDEX

Explanations

No Explanations Found

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

cent

-0.26

Under

-0.26

attr

-0.25

 createState

-0.25

ãĢĤãĢĤ↵↵

-0.25

spirit

-0.25

sei

-0.25

_states

-0.24

 retaining

-0.24

áĥ«

-0.24

POSITIVE LOGITS

åĩ½

0.29

ä¼łç»ŁçļĦ

0.28

},'

0.27

Ä±ÅŁ

0.27

ROUND

0.26

-Compatible

0.26

azole

0.26

ä¸ºäºĨæĽ´å¥½

0.25

 Cheese

0.25

æķĽ

0.25

Activations Density 0.013%

No Known Activations

This feature has no known activations.