INDEX

Explanations

what you need

This neuron detects instances of personal addressing through first- and second-person pronouns (e.g., “I,” “you,” “we,” “my,” “your”).

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

̱

0.62

Zv

0.61

에

0.59

に

0.58

fig

0.58

ื

0.56

'].

0.55

ೕ

0.54

эр

0.54

chromedriver

0.54

POSITIVE LOGITS

 tidak

0.69

 نمی‌

0.69

 sadece

0.67

 yalnızca

0.64

 không

0.64

 hanya

0.63

 weren

0.61

 শুধুমাত্র

0.59

 නො

0.59

 wurden

0.58

Activations Density 0.206%