INDEX

Explanations

open-weights model

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 известный

1.48

 உள்ளார்

1.37

 இருக்கிறார்

1.31

 tecnológico

1.29

 આવ્યો

1.28

 થયો

1.25

 કરવો

1.24

owany

1.23

 poderoso

1.23

 કર્યો

1.21

POSITIVE LOGITS

 αυτές

1.50

它們

1.46

它们

1.45

它们的

1.39

 Elles

1.32

 הן

1.31

 elles

1.30

తున్నాయి

1.30

 españ

1.28

 frescas

1.28

Activations Density 0.009%