INDEX

Explanations

responsible

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Personensuche

-1.05

ⓧ

-0.69

 χρήση

-0.61

RTEX

-0.57

 pinulongan

-0.55

Попис

-0.53

bewerken

-0.53

rsiniz

-0.53

InitVars

-0.52

 становника

-0.52

POSITIVE LOGITS

 killing

0.54

 causing

0.52

 accounts

0.51

 contributing

0.51

NOPQRST

0.49

حياته

0.48

tableFuture

0.48

 increased

0.47

 accounting

0.47

 helping

0.46

Activations Density 0.001%