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

æ³¨æĺİåĩºå¤Ħ

-0.26

dec

-0.26

 briefing

-0.25

ucene

-0.25

mind

-0.25

 Feinstein

-0.25

 paren

-0.25

æµĭè¯Ħ

-0.25

ufen

-0.24

 cháº¡m

-0.24

POSITIVE LOGITS

å¼ĤæĢ§

0.27

bourg

0.27

abled

0.27

bled

0.26

adium

0.26

__).

0.25

srv

0.25

æĬ¼

0.24

olk

0.24

ILD

0.23

Activations Density 0.883%

No Known Activations

This feature has no known activations.