INDEX

Explanations

preferred, preferably, or more preferably

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

-1.82

 this

-1.44

 bílé

-1.33

 mögliche

-1.30

に行きました

-1.29

 остался

-1.27

 lápis

-1.26

Créditos

-1.22

 zůst

-1.22

POSITIVE LOGITS

 также

1.54

 also

1.43

ANCHE

1.41

 canzoni

1.27

 także

1.27

让你

1.26

seys

1.26

また

1.24

тится

1.22

gne

1.21

Activations Density 0.032%