INDEX

Explanations

black, blue, purple

The neuron fires specifically on the word “black,” i.e. it detects mentions of the color black.

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

 रंगों

0.89

 naranja

0.76

 Colors

0.73

 Alemania

0.73

 일본

0.71

 tints

0.71

 Khartoum

0.71

 ピンク

0.69

 রঙিন

0.69

 mixtures

0.68

POSITIVE LOGITS

institution

0.67

orthodox

0.66

 jumpsuit

0.65

を行った

0.63

graduation

0.63

detailed

0.62

hasOwnProperty

0.61

 konten

0.61

 शैक्षणिक

0.61

中所

0.60

Activations Density 0.000%