INDEX

Explanations

hit songs and singles

The neuron fires on numeric tokens and related chart‐ or hit‐count data (e.g. “hit 276,” “charting 13.562,” “hit 264,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

</h3>

-2.28

</sub>

-1.94

最重要

-1.89

</th>

-1.80

῞

-1.77

</strong>

-1.71

౮

-1.62

";

-1.56

alties

-1.56

ERS

-1.54

POSITIVE LOGITS

 potranno

1.85

 kuasa

1.84

 Apenas

1.83

 peringatan

1.83

Both

1.77

 biß

1.69

 Hoje

1.67

ことから

1.66

 assolu

1.65

 suami

1.64

Activations Density 0.014%