INDEX

Explanations

specific terms and entities

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

you

-1.54

 ondu

-1.43

 will

-1.38

 sebuah

-1.38

for

-1.38

rowy

-1.30

vya

-1.30

 rumput

-1.29

 would

-1.27

וּ

-1.25

POSITIVE LOGITS

 ویتامین

1.66

 separar

1.39

 Sünde

1.35

 confirmar

1.34

 excluir

1.32

翻攝

1.31

 ακό

1.29

 cortar

1.29

 кровать

1.27

 during

1.27

Activations Density 0.069%