INDEX

Explanations

operating system

The neuron selectively activates on software‐related gerunds or present‐participle terms—especially “operating,” “processing,” and “programming”—as they appear in technical contexts.

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

to

-1.91

ᜀ

-1.59

of

-1.51

in

-1.50

瘜

-1.47

栒

-1.38



-1.36

いただけます

-1.32

 réputation

-1.31

蟥

-1.30

POSITIVE LOGITS

and

2.14

Alguns

1.49

 this

1.47

 various

1.46

 settore

1.38

 During

1.36

␻

1.36

 different

1.34

Após

1.34

Apesar

1.34

Activations Density 0.491%