INDEX

Explanations

relates to

The neuron primarily picks out numeric values—especially decimal numbers and measurements—in technical or patent‐style text.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 when

-1.92

and

-1.61

;*/

-1.38

فوائد

-1.37

'':

-1.36

 '*',

-1.35

 without

-1.33

are

-1.32

ר

-1.30

ଞ

-1.27

POSITIVE LOGITS

にぎり

1.52

 unceasing

1.38

撂

1.37

就这样

1.33

 zoeken

1.31

ولايات

1.29

 monstrous

1.27

឵

1.27

 brisk

1.27

où

1.27

Activations Density 0.754%