INDEX

Explanations

countries and regions

The neuron fires most strongly on numeric tokens and numeric identifiers (e.g. version numbers, quantities, IDs).

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

-1.02

-0.87

 maraming

-0.80

больше

-0.80

))));

-0.79

わけで

-0.79

ख्या

-0.77

 Eingang

-0.76

ptăm

-0.76

ホルダー

-0.75

POSITIVE LOGITS

occas

1.02

 отдель

0.91

单独

0.88

 minY

0.87

manuelle

0.86

izien

0.84

专

0.83

があります

0.82

 negligible

0.81

actividad

0.80

Activations Density 0.029%