INDEX

Explanations

Tennis match results and scores

The neuron primarily detects decimal‐formatted numbers (numerical values containing a point, e.g., “15.125,” “66.5,” 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

``,

-1.66

 ktorí

-1.65

 ktorá

-1.63

 которое

-1.59

 ktoré

-1.54

璠

-1.53

 majú

-1.52

膺

-1.51

 която

-1.50

POSITIVE LOGITS

 dunklen

1.45

1.40

1.38

Without

1.34

With

1.32

1.29

 Empfindung

1.28

dö

1.27

Taking

1.27

män

1.26

Activations Density 0.010%