INDEX

Explanations

class names and identifiers

The neuron activates on numeric literals with decimal points.

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

會有

-1.46

PRESIDENT

-1.41

probably

-1.40

 siguran

-1.37

Эти

-1.37

Methode

-1.35

 unmistak

-1.33

 результаты

-1.32

 线条

-1.31

 certamente

-1.29

POSITIVE LOGITS

 this

1.75

1.72

to

1.70

 from

1.67

was

1.66

</em>

1.64

on

1.55

for

1.52

 because

1.51

</b>

1.50

Activations Density 0.002%