INDEX

Explanations

starting actions or states

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

 also

-2.48

 follows

-1.43

),

-1.38

 також

-1.30

 prezzi

-1.28

also

-1.27

 też

-1.27

되는

-1.27

 again

-1.26

 verbindet

-1.24

POSITIVE LOGITS

нование

1.52

 Mungkin

1.49

毬

1.45

 ممکن

1.44

 Parece

1.41

 Frases

1.37

 jetzt

1.35

Få

1.35

 Kasus

1.30

鈎

1.30

Activations Density 0.065%