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

çĥĽ

-0.28

è¥Ħ

-0.26

tip

-0.26

 AltÄ±n

-0.25

ä¸Ĭåįĩ

-0.25

 defaultProps

-0.24

ç®¨

-0.24

hel

-0.24

æ¼³

-0.23

/contact

-0.23

POSITIVE LOGITS

istan

0.28

sut

0.26

oms

0.26

ÑĥÑĩÐ°

0.25

å¿ı

0.25

ocate

0.24

angu

0.24

å¤§ä¼Ĺ

0.23

ä¾ĽåºĶéĵ¾

0.23

ç§ĳ

0.23

Activations Density 0.031%

No Known Activations

This feature has no known activations.