INDEX

Explanations

This neuron spotlights standalone topic-label words that serve as section headings or list titles (e.g., “Symptoms,” “Plugins,” “Advantages”).

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

 والذي

0.91

పడుతుంది

0.81

 বলেছিল

0.75

 தெரிவித்துள்ளது

0.73

 quale

0.73

 وهو

0.70

விட்டது

0.69

 कहता

0.69

 বলেছে

0.68

 βρίσκεται

0.67

POSITIVE LOGITS

are

2.95

 were

2.72

 jsou

2.66

 ovat

2.66

są

2.53

 đều

2.52

 são

2.44

都是

2.44

 هستند

2.39

 aren

2.34

Activations Density 0.713%