INDEX

Explanations

mathematical expressions and definitions

This neuron activates on numeric tokens (numbers or numeric literals) in the text.

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

-2.11

-1.70

៚

-1.65

-1.64

-1.63

は

-1.57

 that

-1.55

<bos>

-1.51

!（

-1.43

Pgina

-1.37

POSITIVE LOGITS

</strong>

2.36

and

1.93

 certamente

1.63

れます

1.59

 květ

1.57

簠

1.55

derdag

1.55

during

1.53

更快

1.52

 diametro

1.51

Activations Density 0.002%