INDEX

Explanations

magic and spells

The neuron activates on words and phrases evoking magical or fantasy elements (e.g. spells, magic, mythical creatures).

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

 sashimi

0.67

🦍

0.64

 नक्स

0.61

 मुना

0.59

 Goliath

0.58

 carnage

0.57

 Infantry

0.57

 giraffe

0.57

🚁

0.55

 Padukone

0.54

POSITIVE LOGITS

魔法

1.79

 witchcraft

1.74

 magick

1.74

 magic

1.73

 magical

1.72

 magia

1.67

 маги

1.66

 wizard

1.60

🧙

1.59

 wizards

1.59

Activations Density 0.482%