INDEX

Explanations

tree

The neuron is specifically triggered by mentions of trees (the word “tree”/“trees”).

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

 straw

0.57

 หัว

0.54

 المض

0.54

 لح

0.52

 bandes

0.52

Straw

0.52

 makarna

0.52

 земле

0.51

ライス

0.50

 تہ

0.50

POSITIVE LOGITS

 tree

2.83

树

2.80

 trees

2.78

 Trees

2.73

Trees

2.70

樹

2.69

 Tree

2.67

Tree

2.63

 درخت

2.56

tree

2.55

Activations Density 0.456%