INDEX

Explanations

`required` or `|` following code syntax

This neuron detects numeric literal tokens (numbers) in the code or data.

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

RE

-1.48

RA

-1.38

用

-1.34

want

-1.33

actéristi

-1.31

atively

-1.30

两个

-1.29

Ꮸ

-1.29

indicate

-1.27

 $^{\

-1.27

POSITIVE LOGITS

 jednak

1.61

is

1.55

 which

1.50

uckland

1.46

vegli

1.43

 similarly

1.42

 vaše

1.41

stacle

1.41

if

1.40

 devido

1.40

Activations Density 0.013%