INDEX

Explanations

mathematical terms and notation

The neuron activates on number tokens (integer or decimal) in mathematical or technical expressions—that is, it detects numeric literals in equations.

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

ჱ

-1.61

なんとなく

-1.59

这事

-1.58

쑹

-1.56

 remé

-1.56

鰯

-1.52

 การ

-1.51

 économ

-1.51

 perguntou

-1.51

いっ

-1.50

POSITIVE LOGITS

and

4.00

but

2.02

 with

1.94

or

1.86

in

1.67

as

1.64

 which

1.63

 Wartawan

1.55

 while

1.55

 what

1.48

Activations Density 0.051%