INDEX

Explanations

The neuron primarily detects numeric tokens (especially decimal numbers).

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

 tambien

0.73

 nhưng

0.67

 pero

0.66

Then

0.66

 אבל

0.65

然后

0.64

 lakini

0.64

 لكن

0.63

 Then

0.63

 પણ

0.63

POSITIVE LOGITS

 namely

0.48

 우선

0.48

 nämlich

0.47

 posits

0.46

 nominally

0.46

によると

0.45

 totiž

0.45

 먼저

0.45

 Öncelikle

0.45

 Согласно

0.44

Activations Density 0.000%