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

åĬ¨éĿĻ

-0.27

-tier

-0.27

éĿ¡

-0.26

åĩłä¸ª

-0.26

æĮĩä»¤

-0.25

 Allocate

-0.25

ç§»åĪ°

-0.25

ä¸Ģå¥Ĺ

-0.24

 acceler

-0.24

quila

-0.24

POSITIVE LOGITS

ivic

0.28

ä»Ĺ

0.27

iod

0.27

hard

0.26

@a

0.26

 Bread

0.26

Hard

0.26

odd

0.24

UnitOfWork

0.24

 Hard

0.24

Activations Density 0.023%

No Known Activations

This feature has no known activations.