INDEX

Explanations

The neuron is triggered by numeric expressions (numbers, dates, measurements) in the text.

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

बानी

0.60

DOMContentLoaded

0.57

 Aces

0.54

agati

0.54

ágina

0.53

udges

0.53

 लाग

0.53

ifte

0.51

 বলিউ

0.51

 লোপ

0.51

POSITIVE LOGITS

埴

0.58

<u>

0.54

explicit

0.52

 measurements

0.51

र्वाद

0.51

優しい

0.50

measurement

0.50

 மீட்ட

0.50

Dónde

0.48

ધાર

0.47

Activations Density 0.028%