INDEX

Explanations

Plaintiff versus Defendant

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

цтва

-1.41

four

-1.35

倏

-1.34

-1.33

}{}

-1.30

 utiliser

-1.27

-1.26

скі

-1.26

婆婆

-1.24

only

-1.23

POSITIVE LOGITS

 Both

1.50

maları

1.49

他們

1.46

 сначала

1.44

Their

1.38

主要是

1.35

 methods

1.32

 monasterio

1.30

 Проци

1.30

则是

1.26

Activations Density 0.071%