INDEX

Explanations

used as a function

The neuron activates on the word “as” when it’s used to introduce roles, functions, or comparisons.

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

-1.71

 SqlParameter

-1.42

 SqlCommand

-1.41

mintaan

-1.38

ご了承

-1.38

-1.36

在他

-1.29

 عرو

-1.26

 DialogResult

-1.26

 室内

-1.26

POSITIVE LOGITS

</em>

1.63

 راحتی

1.35

idered

1.33

 eikä

1.32

:");

1.32

 obiettivo

1.30

 mahdol

1.27

脸

1.25

‘

1.23

JEL

1.23

Activations Density 0.004%