INDEX

Explanations

addresses and locations

The neuron activates on numeric tokens (especially standalone numbers, dates, measurements, or decimals).

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

 OGSÅ

-1.90



-1.66

⁛

-1.64

 same

-1.59

should

-1.51

鷙

-1.48

of

-1.46

 beeind

-1.42

駸

-1.42

今年も

-1.41

POSITIVE LOGITS

</b>

2.05

</em>

2.00

</sup>

1.78

</s>

1.70

 möchten

1.53

 siempre

1.52

</code>

1.46

CHREIB

1.43

 quería

1.43

UAWEI

1.41

Activations Density 0.075%