INDEX

Explanations

aetheria

This neuron responds to capitalized names and abbreviations—i.e. proper nouns, acronyms, and similar named entities.

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

 decriminal

0.70

 tradeoffs

0.69

 preventative

0.67

 aforementioned

0.66

 quarterbacks

0.64

 precautionary

0.63

 retroactive

0.63

 preseason

0.62

 statesman

0.62

 defenseman

0.62

POSITIVE LOGITS

 Ха

0.73

Ба

0.73

С

0.72

 Ба

0.70

ار

0.69

－

0.68

ﺴ

0.68

 Ку

0.67

0.66

Activations Density 1.419%