INDEX

Explanations

trading, forex, model, project regulation

The neuron detects marketing‐style product descriptions—that is, phrases highlighting features, quality, and benefits (e.g. “stylish and comfortable,” “high quality non-abrasive mesh,” “front pivot mechanism”).

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

 nonprofits

1.36

 nonprofit

1.29

 美国

1.20

 kvůli

1.09

 Spokane

1.07

 ACLU

1.07

 콘텐츠

1.06

 Když

1.05

 molecules

1.05

 것처럼

1.04

POSITIVE LOGITS

Whilst

1.71

 Whilst

1.43

 Amongst

1.26

 whilst

1.23

 enquiry

1.23

 हेतु

1.21

 ಹಾಗೂ

1.21

にて

1.20

更に

1.20

ครับ

1.19

Activations Density 1.893%