INDEX

Explanations

as in the case

as is/was [descriptor]

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

0.47

 stärker

0.46

н

0.39

 econ

0.38

ంతా

0.38

 सुनकर

0.37

さんが

0.37

Toggler

0.37

SwingConstants

0.36

kao

0.36

POSITIVE LOGITS

 경우

0.55

 випадку

0.55

 costumbre

0.52

SOP

0.52

 περίπτωση

0.52

ゼ

0.51

 경우는

0.50

の場合

0.50

є

0.49

 場合

0.49

Activations Density 0.063%