INDEX

Explanations

starts with a

checking string prefixes

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

[{{

0.47

 fans

0.47

 idol

0.46

 altera

0.45

..........

0.45

 eldest

0.43

 conspiring

0.43

)^{\

0.42

гү

0.42

 singing

0.41

POSITIVE LOGITS

OWA

0.53

 Kombination

0.50

گون

0.50

 Kedua

0.49

بد

0.49

 Amerikaanse

0.48

 Schalt

0.48

emeen

0.47

الصفحه

0.45

噦

0.45

Activations Density 0.000%