INDEX

Explanations

listing numbers and measurements

The neuron activates on numeric values (quantitative measurements and percentages) 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

 επίσης

-1.77

๗

-1.48

也是

-1.39

 også

-1.39

 ALSO

-1.34

 भी

-1.32

dă

-1.28

бо

-1.27

 Quelles

-1.26

도

-1.26

POSITIVE LOGITS

 they

1.92

–

1.53

—

1.51

 various

1.51

 their

1.47

 that

1.44

 some

1.36

 more

1.35

1.30

 other

1.30

Activations Density 0.155%