INDEX

Explanations

involvement and participation

The neuron detects direct appeals to or instructions for the reader—second‐person (“you,” “your”) and participation/involvement language.

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

器

0.51

 folos

0.51

캠

0.51

開店

0.50

 zásob

0.50

udd

0.49

 crackers

0.49

 библиоте

0.48

 rassembl

0.48

 reunions

0.47

POSITIVE LOGITS

 involvement

3.25

 Involvement

3.09

 participation

2.80

参与

2.78

參與

2.69

 Participation

2.67

Participation

2.64

participation

2.63

 involved

2.55

 participación

2.50

Activations Density 0.834%