INDEX

Explanations

include some

The neuron fires on marketing-style benefit statements describing advantages or features of a product or service.

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

 primarily

0.89

 first

0.77

首先

0.77

 justru

0.77

 głównie

0.75

first

0.75

 pertama

0.73

0.71

 only

0.71

 esclusivamente

0.70

POSITIVE LOGITS

 अन्य

1.27

 અન્ય

1.20

其他的

1.16

 다른

1.13

 других

1.09

的其他

1.08

 ಇತರ

1.07

 другими

1.06

 outras

1.03

 outros

1.02

Activations Density 0.720%