INDEX

Explanations

mathematical expressions with variables

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

 wesentlichen

-1.74

 offiziellen

-1.55

 zarówno

-1.45

くださ

-1.45

 descrito

-1.39

 विश्वसनीयता

-1.39

下さ

-1.38

咄

-1.37

دکتر

-1.36

くださった

-1.36

POSITIVE LOGITS

 cromado

1.53

‘‘

1.52

Anyways

1.52

🕗

1.52

 darbu

1.51

SPECIAL

1.45

 encaje

1.44

 fjor

1.44

Mostly

1.41

highly

1.41

Activations Density 0.001%