INDEX

Explanations

lu

The neuron fires on longer, formal technical or legal-style words (e.g. collection, nature, identification, diagnosis) used in policy and specialized contexts.

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

 είναι

0.84

 例如

0.73

 explored

0.67

 plays

0.66

 hanging

0.64

itabbo

0.62

例如

0.59

itabbam

0.57

 রাসেল

0.56

 Kecamatan

0.55

POSITIVE LOGITS

and

2.17

และ

1.95

 અને

1.87

и

1.86

và

1.86

 και

1.81

ও

1.79

 आणि

1.77

 और

1.72

 ਅਤੇ

1.72

Activations Density 0.000%