INDEX

Explanations

The neuron fires on the names of categories or tags (typically capitalized labels) in post metadata.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 सम्पूर्ण

0.57

අ

0.57

 কারনে

0.56

англ

0.55

 महाविद्या

0.54

sic

0.54

您可以

0.54

الأ

0.54

 దీంతో

0.54

 न्यायालयाने

0.53

POSITIVE LOGITS

،

2.91

、

2.88

2.78

၊

2.58

፣

2.52

，

2.27

-,

2.09

、

1.98

,\

1.95

‌,

1.90

Activations Density 0.634%