INDEX

Explanations

higher values or occurrences

The neuron activates on intensity or degree words—particularly “high” and other modifiers (e.g., “very,” “higher,” “low”) that signal magnitude.

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

zwischen

-1.42

syk

-1.41

Китай

-1.38

Современ

-1.34

طراحی

-1.33

 mancha

-1.32

 haki

-1.30

 μας

-1.29

uchungen

-1.28

 rusa

-1.27

POSITIVE LOGITS

 that

2.14

to

1.33

 probabilidad

1.31

 значит

1.31

 что

1.30

you

1.27

1.25

 help

1.23

 These

1.23

Activations Density 0.045%