INDEX

Explanations

AF constants and class names

The neuron fires on numeric literals (numbers, often as part of code constants) in source code.

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

it

-1.70

 beträ

-1.54

溷

-1.48

 erwor

-1.44

 vicinity

-1.43

 extent

-1.40

/¯

-1.35

ೣ

-1.33

 respective

-1.32

の影響

-1.32

POSITIVE LOGITS

</h2>

1.70

There

1.63

From

1.53

Since

1.47

Having

1.46

Unlike

1.34

Using

1.33

Before

1.31

1.30

Without

1.30

Activations Density 0.013%