INDEX

Explanations

Components II ends Order dreams Property data When

The neuron fires on numeric tokens—particularly floating‐point or decimal numbers (e.g. volume references, measurements).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ро

0.89

ات

0.87

ς

0.83

0.82

س

0.76

Subview

0.76

兵庫

0.75

חו

0.75

NewLabel

0.72

,//

0.71

POSITIVE LOGITS

 personalizados

0.94

 masks

0.93

 tabulated

0.91

 Pupils

0.88

 Masks

0.85

 Wetlands

0.85

 frescas

0.84

 Wildcats

0.83

 puzzling

0.82

hrs

0.81

Activations Density 0.001%