INDEX

Explanations

prefix, order, post/pre, exposure

This neuron detects technical terms related to ordering or traversal—e.g. words like “sorted,” “order,” “pre-order,” “traversal,” “prefix,” or “postfix.”

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

geſ

-1.67

 Alquiler

-1.59

 vostri

-1.54

᷆

-1.50

 médicas

-1.45

 deportivas

-1.44

phology

-1.41

 filmu

-1.40

matician

-1.39

ندم

-1.39

POSITIVE LOGITS

1.63

1.51

</em>

1.48

1.46

—

1.41

of

1.31

1.30

様々

1.27

Activations Density 0.017%