INDEX

Explanations

"our" and surrounding phrases

The neuron detects numeric time and ordinal expressions—especially mentions of years (e.g. “years ago”), dates, ordinals (e.g. “39th”), and similar number-based factual references.

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

了一会儿

-1.38

 deux

-1.37

看完

-1.34

 puedes

-1.29

 ellos

-1.29

 Autorin

-1.27

 principalement

-1.26

 regionales

-1.24

粉色

-1.23

acán

-1.22

POSITIVE LOGITS

our

3.80

we

2.50

 your

1.95

我们

1.80

my

1.78

you

1.63

我們

1.55

1.52

เรา

1.52

我们的

1.49

Activations Density 0.007%