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

æķ£

-0.28

ç¦ıåĪ©

-0.27

ocations

-0.27

trial

-0.26

 twins

-0.25

DU

-0.25

æĹ¶æĬ¥

-0.25

 sweep

-0.25

 blush

-0.25

tw

-0.24

POSITIVE LOGITS

TES

0.26

Ã¶rÃ¼

0.25

 Ä°s

0.25

 ÑĥÐºÑĢÐ°

0.25

[âĢ¦

0.25

æ·Ģç²ī

0.24

ä¼ªéĢł

0.24

æ·Ģ

0.24

allet

0.24

(Of

0.23

Activations Density 0.000%

No Known Activations

This feature has no known activations.