INDEX

Explanations

share this article

The neuron strongly activates on numerical figures and quantifiable data (digits, percentages, scores, counts).

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

 vérification

-1.63

 ſont

-1.50

かも知

-1.43

莇

-1.41

 mici

-1.39

earn

-1.36

see

-1.34

 voyez

-1.34

 工事

-1.33

 envíos

-1.32

POSITIVE LOGITS

or

1.68

if

1.61

 when

1.56

1.55

1.48

wir

1.45

at

1.43

_{

1.41

 Perhaps

1.37

 Usually

1.36

Activations Density 0.028%