INDEX

Explanations

namelists

This neuron detects the frequent “ul” letter sequence found in Arabic-style personal names (e.g. “Khfan ul”, “Ta hir ul”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

一定的

-1.62

''

-1.59

-1.54

of

-1.44

-1.42

 compositeur

-1.39

 strenuous

-1.38

like

-1.38

 sencilla

-1.37

POSITIVE LOGITS

</sup>

2.03

 posibil

1.83

,’

1.73

我們

1.70

ྒྱ

1.64

 sorella

1.63

 This

1.57

 Tentang

1.56

 illustrated

1.54

 możemy

1.54

Activations Density 0.007%