INDEX

Explanations

attributed sources and findings

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

噘

-1.33

 nawet

-1.31

 karena

-1.30

 contará

-1.30

↵

-1.27

ұл

-1.27

 nascita

-1.26

kennen

-1.24

 recibo

-1.23

 tecnologías

-1.23

POSITIVE LOGITS

for

1.83

as

1.54

 Also

1.48

 also

1.48

 during

1.45

 both

1.43

by

1.38

one

1.33

 always

1.30

 usually

1.27

Activations Density 0.106%