INDEX

Explanations

Pakistan and Pakistani topics

The main thing this neuron does is detect occurrences of the word “blasphemy” (and its morphological fragments).

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

 Bartholom

0.61

 Girona

0.61

 Noyes

0.61

 Voit

0.60

 ذریعہ

0.59

 запол

0.57

 Loire

0.56

 suelos

0.56

हिणी

0.55

 Reims

0.55

POSITIVE LOGITS

 Pakistan

1.95

Pakistan

1.92

 Pakistani

1.82

 پاکستان

1.77

 Islamabad

1.75

 Karachi

1.74

 pakistan

1.70

 पाकिस्तान

1.68

 Lahore

1.60

 पाकिस्तानी

1.59

Activations Density 0.364%